Abstract:The research and development of advanced agricultural machinery equipment in hilly and mountainous areas is one of the hot and difficult problems in the development of agricultural machinery equipment in China. At present, the agricultural production in hilly and mountainous areas still has the practical problem of “no available machinery, no good machinery”, and the research and development of advanced suitable agricultural equipment lacks the necessary theoretical support. The research status of hillside agricultural machinery equipment at home and abroad, especially hillside tractor, hillside agricultural machinery leveling technology, hillside tractor driving and power system, hillside machinery and tools and their operating performance were summarized, and the research progress of soil compaction and soil tillage erosion in slope operation by agricultural machinery was analyzed. The advanced computer aided methods for the design of hillside agricultural machinery and the study of the interaction mechanism between hillside agricultural machinery and soil were summarized. Finally, the research focus and development direction of agricultural machinery equipment and slope operation technology in hilly and mountainous areas were prospected as follows: the interaction mechanism between agricultural machinery and slope soil and crop (nutrient); the overall design of hillside agricultural machinery deeply integrated with agronomy and slope operation scene; efficient and reliable attitude adjustment mechanism and control strategy of hillside machinery; hillside farm machinery efficient transmission and easy steering drive; and intelligent monitoring and accurate autonomous navigation of mountain farm machinery operation. This review can provide reference for efficient and high quality production of agriculture in hilly and mountainous areas.

Abstract:To meet the demand for unmanned operations in rice and wheat harvesting, an algorithm using LiDAR to detect the rice and wheat harvest boundary was proposed, and the automatic alignment of the harvest boundary was realized by connecting the unmanned control system. Firstly, the algorithm delimited the angle range of interest for the collected harvesting outline point cloud, converted the measured data from polar coordinates to three-dimensional rectangular coordinates according to the installation height and position of the LiDAR, and corrected the measured point cloud by fusing the LiDAR attitude measured by the gyroscope. The noise and non-crop contour points in the point cloud were filtered by median filtering and Z-direction threshold filtering. The accuracy of K-means clustering and Z-direction central difference method for detecting harvest boundary was compared, and the error analysis was carried out. The sensing system was developed and the CAN communication protocol of sensing and control was established. Point tracking strategy was adopted to automatically control the boundary points detected in real time. The automatic alignment accuracy detection method of rice and wheat harvest boundary was analyzed and studied. In June 2022, the experiment of harvesting boundary detection and automatic alignment control system was carried out at Xiaotangshan National Precision Agriculture Demonstration Base in Beijing. The data were collected and analyzed by using data annotation and GPS pointing respectively. The experiment showed that the average horizontal error of harvesting boundary detection based on K-means clustering was 22.24cm, the average horizontal error of the Z-direction central difference method was 1.48cm, and the Z-direction central difference method was superior to the K-means clustering method. Therefore, the Z-direction central difference method was used for automatic alignment control experiment. The average value of lateral deviation of automatic alignment control system was 9.18cm, and the standard deviation was 2.48cm. The system can be used for unmanned rice and wheat harvesting.

Abstract:In order to reduce the cumulative drift error of the orchard robot in simultaneous localization and mapping(SLAM), a loop closure detection algorithm was proposed based on density binary pattern(DBP). The LiDAR scanning was divided into eight-bit binary bins along the vertical height direction. If the number of point clouds in the bin exceeded five, it was considered a valid scan, and the bin value was set to be 1, otherwise 0. Further, the eight-bit data were projected to construct the DBP descriptor. The DBP descriptor contained point cloud density and height information and had a significant distinguishing effect on tall fruit trees and low shrubs. A two-stage search algorithm was utilized to ensure the task real-time requirements in the large-scale orchard. Firstly, to extract a low-dimensional ring factor vector of DBP, the K-nearest neighbor candidate loop closure frames were quickly found in the K-dimensional tree(KD-Tree), which was constructed by the ring factors. The maximum similarity between the candidates and the query frame was obtained. If the distance threshold condition was met, the candidate frame was considered an effective target loop closure. The experiment was carried out in three orchards of different scales. In the orchard scene with multiple loop closure events, the root mean square error and standard deviation of the DBP-LeGO-LOAM trajectory were 0.24m and 0.09m, compared with the LeGO-LOAM algorithm which had been reduced 81% and 91% respectively. It provided an effective solution for improving the mapping and localization accuracy of orchard robots.

Abstract:In order to reduce the labor intensity of manual topping, the environmental pollution of chemical topping and the over-topping of the traditional “one-size-fits-all” topping mechanism, a double disc clamping topping mechanism was designed by analyzing the manual topping process. Based on machine vision, a single-line prototype of the topping device was designed to realize the automatic control of the whole process of cotton topping. It was mainly composed of topping mechanism, visual inspection mechanism, motion mechanism, cotton top recognition and control system. Based on cotton field research, structural calculation and pre-test, the overall structure and key component dimensions of the topping device were determined. Combined with the research basic and practical application of visual recognition, the YOLO v3 algorithm was selected to build the cotton top recognition and control system, realize the recognition and positioning of cotton top, and complete the motion control of the topping mechanism. Taking cotton in the topping stage as the research object, the cotton top recognition test, the topping mechanism performance test and the field comprehensive test were carried out. The results showed that the average recognition rate of cotton top recognition test was 93%;the average topping rate of the performance test of the topping mechanism was 94.67%;the average recognition rate and average topping rate in the field test was 85.33%, and the average topping rate was 78.22%. The research result can provide reference for the precise and intelligent research of cotton topping.

Abstract:A rotary-cut micro-ridge seedbed preparation device (RMSD) was designed to facilitate rapeseed direct seeding in the mid-lower reaches of the Yangtze River region, where continuous rainfall and sunny days can cause significant fluctuations in soil moisture and affect seedling growth. The RMSD employs rolling boards that passively rotate and stack micro-ridges by cutting into the rotary-tilled soil as the machine moves forward, with grooves connecting to furrows on both sides. The geometric size of the micro-ridge was defined according to the row spacing configuration and root growth characteristics of direct rapeseed seeding, combined with soil water and ridge cultivation heat characteristics. Based on kinematics, the rolling board endpoint trace and the rolling board envelope were analyzed to determine the qualitative parameters in the structure terms. The results were that the number of rolling boards was 6~10, and the angle range of the rolling board was 0°~28°for further study. The DEM-MBD coupling simulation was conducted with the single factor test and the orthogonal rotation regression test to investigate and optimize the parameters of the RMSD (rolling board number, rolling board angle, and forward speed). The optimization results were as follows: when the rolling board of the RMSD was 6 and 8, respectively, and the rolling board angle, rotary cutting depth, and efficient groove depth were 28.00°, 100mm, 83.59mm and 26.50°, 92mm and 64.26mm, respectively. The field test used the Trimble TX8 3D laser scanner to reconstruct the micro-ridge surface after RMSD operation and compared it with the simulation results. The field test showed the maximum deviations of actual depth and micro-ridge distance were 8.25%, and the qualified rates of micro-ridge were 100% and 90%, respectively, under the optimal parameter combination of 6 and 8 rolling boards. The research result can provide a reference for the structural improvement of the rapeseed bed preparation device.

Abstract:In view of the technical problems of low sowing quality caused by re-sowing and serious leakage of seeds when the existing air-absorbing seed meter was used for high speed sowing operation, a combination of repellent and guide slot auxiliary seed attachment air-absorbing peanut high-speed precision seed meter was designed. The design of the seed discharge performance under high-speed operation was carried out. The theoretical modeling analysis verified the rationality of structure design of the seed discharge disc and initially completed the determination of key parameters. In order to further optimize the size of seed churning recess and seed pick-up slot, the influence of key parameters on seed transport was analyzed with the help of discrete element simulation software, and a two-factor, five-level quadratic orthogonal rotational combination test and regression analysis were conducted to further optimize the structural parameters. The optimal combination of the following parameters was obtained: the depth of the seed churning recess was 3mm, the radius of base circle of churning recess was 70mm, the distance between left and right edges of seed picking slot was 24.0mm, the distance between lower edge of seed picking slot was 19.1mm, the depth of seed picking slot was 10.5mm, and the distance from outer circumference of seeding disc to rear end of seed picking slot was 24.0mm. When the wind pressure was -6kPa and the operating speed was in the range of 6~12km/h, the particle spacing qualification index was no less than 93.33%, the reseeding index was no more than 3.52%, the omission index was no more than 4.02% and the breakage index was no more than 0.32%, which had good operating performance.

Abstract:When transplanting on the film, the vegetable transplanter can only adjust the plant spacing in a small range to ensure a smaller planting hole opening, if the plant spacing is too large, it will cause the planters to move forward and tear the film, if the plant spacing is too small, it will cause the planters to move backward with the film, these will make the planting hole opening too large, and will affect the later growth of seedlings, for this reason a set of reciprocating duckbill planting device with steplessly adjustable planting static trajectory was designed. A set of planting mechanism with steplessly adjustable planting trajectory and its planting method were proposed based on the analysis of the structural characteristics of the planting device, the working process and the principle of adjusting the size parameters of the components and the installation parameters to realize the planting of smaller holes with different plant spacings. By establishing a veridical kinematic model of the mechanism, developing visual aided design software, analyzing the influence of each parameter on the planting characteristics, and using the stepwise approximation method to determine a set of parameters that met the requirements as follows：L1=35mm, L2=350mm, L3=70mm, L5=280mm, dD=358mm, Φ4=15°, xB=20mm. The position of the longitudinal coordinate yB of the pivot point B corresponding to every 50mm interval between the plants was marked. The duckbill planting device was designed and modelled in 3D according to the combination of mechanism parameters, and kinematic simulations were carried out to verify the feasibility of the duckbill planting device. Laboratory prototype tests were carried out and the results showed that the planting pass rate was greater than 90% at plant spacings between 100mm and 600mm, the coefficient of variation of plant spacing was less than 6% and the hole size was less than 100mm, with an average hole size of 52.26mm at plant spacing of 100mm.

Abstract:Potted seedlings need to be transplanted from the plug tray to the cultivation slot in greenhouses, while the inefficient manual operation limits the large-scale production. In order to achieve efficient and high-quality automatic transplanting operation, a kind of row picking end-effectors with plug-in claws was designed for high-speed transplanting operation of greenhouse pot seedlings. The stress and the deformation of the transplant effector during taking and releasing seedlings were analyzed. Combined with ADAMS rigid-flexible coupling simulation, the optimization design of the transplant effector was carried out to determine the fitting curve of the claw tip point and the gripping trajectory for the pot seedling. Taking the main factors affecting the success rate of taking and releasing seedlings, such as grasping seedling depth, substrate moisture content of matrix, lifting speed and horizontal speed, as the variables, the orthogonal test was carried out, and the optimal parameter combination was determined. Based on the results of examination，it can be found that the best parameter combination was the grasping seedling depth of 48mm, moisture content of matrix of 69.9%, lifting speed of 0.24m/s and horizontal speed of 0.35m/s. Under the optimal condition, the success rate of taking and releasing seedlings was 97.9%, and the efficiency reached 10322 plants/h, which can meet the requirements of high-speed and efficient transplanting. The research result can provide a reference for the domestic development of high-speed pot seedlings transplanter in greenhouses.

Abstract:In view of lack of accurate models for discrete element simulation in the current research and development process of forage harvesting, crushing, processing equipment transportation and cutting, the alfalfa budding stem with high moisture content and complex physical and mechanical characteristics was taken as the research object. With the aid of EDEM simulation software, the physical parameters and bonding parameters were calibrated based on Hertz-Mindlin (no slip) and Hertz-Mindlin with bonding contact models respectively. Based on angle of repose and shear test, physical parameters such as Poisson’s ratio, shear modulus, impact recovery coefficient, static friction coefficient, rolling friction coefficient and bonding parameters such as normal contact stiffness, tangential contact stiffness, critical normal stress, critical tangential stress and bonding radius of alfalfa stem were determined through Plackett-Burman test, the Steepest ascent test and Box-Behnken test. The relative error between the simulated angle of repose and the physical angle of repose was 0.52%. The results showed that the relative error between the simulated shear failure force and the physical test simulation failure force was 0.86%, which indicated that the calibrated parameters can truly reflect the physical and mechanical characteristics of alfalfa stems at the budding stage. It provided a reliable model and parameter calibration method for the discrete element simulation in the research and development process of forage machinery, and also provided a reference for the research and development, optimization design and power matching of the conveying and cutting mechanisms of forage harvesting and crushing processing equipment.

Abstract:In response to the existing cotton stalk harvesting machinery has high breakage rate, high leakage rate, and the need of work on rows and other problems, a clamping-roller cotton stalk pulling device was designed. The device was mainly composed of cotton stalk pulling mechanism and cotton stalk conveying mechanism, and the structural parameters and working parameters of each component were determined through the kinematic and dynamics analysis of the operation process of the cotton stalk pulling mechanism. In order to verify the reliability and operational performance of the cotton stalk pulling device, a three-factors, three-levels quadratic regression orthogonal test was conducted with the forward speed of machine, the speed of upper stalk pulling roller, the ratio of forward speed of the machine to the linear speed of the stalk plucking wheel (referred to as the speed ratio) as the test factors, and the stalk breakage rate and the leakage rate as the test indexes. A regression model was established to analyze the influence of each factor on the operational performance of the cotton stalk pulling device, and parameter optimization and test verification were carried out. The test results showed that the factors affecting the stalk breakage rate in the main order was upper stalk roller speed, machine forward speed and speed ratio;the factors affecting the stalk leakage rate in the main order was speed ratio, machine forward speed and upper stalk roller speed. The optimized working parameters was as follows: machine forward speed of 0.60m/s, upper stalk roller speed of 46r/min and speed ratio of 0.50 for field test, the combination of these parameters was tested in the field, and the stalk breakage rate was 3.68%, and leakage rate was 5.19%, the relative error with the theoretical optimized value was not more than 5%, the research results can provide reference for the design of cotton stalk pulling device.

Abstract:Aiming at the problem of lack of accurate bonding parameters in the discrete element simulation model of rape shoots stems, which is closely related to the key links of mechanized harvesting of rape shoots stems, such as cutting, transport by clamping, and baling, the stems of mechanized clamping section of rape shoots in the double low rape harvesting period of “oil and vegetable dual-use” were used as research objects. Using EDEM simulation software, a method to construct a discrete element simulation model of double-layer bonding of the middle clamping section stems of the rape shoots was proposed by using three-axis spatial coordinates method. Design-Expert software was used to design the Plackett-Burman test, the steepest climb test and the Box-Behnken test to complete the calibration of the simulation bonding parameters of the clamping section stems of rape shoots. The shear and radial compression models were constructed by using the optimized solutions of calibrated parameters, and the corresponding simulation tests were carried out. The model parameters were further optimized by comparative analysis with the physical test. The results showed that the normal/tangential contact stiffness of inner core to inner core, the normal/tangential contact stiffness of epidermis to inner core, and the normal contact stiffness of epidermis to epidermis had significant effects on the mechanical properties of stems, and the results were 1.94×107N/m, 9.56×108N/m and 6.28×109N/m, respectively. The relative error between simulated values and measured values of all mechanical models was not greater than 3%. And the force change trend of the stem was basically the same, indicating that the optimized calibration parameters were feasible and accurate, and the constructed double-layer bonding discrete method model of stems of rape shoots characterized the difference in mechanical characteristics of its internal structure, which can provide a model basis for the numerical simulation of the related system of stem of rape shoots.

Abstract:In order to improve the accuracy and reliability of using the discrete element method to guide the design and optimization of banana straw crushing and returning equipment, Hertz-Mindlin with bonding contact model was used to establish the discrete element bonding model of banana straw and carry out parameter calibration. The impact recovery test, static friction test, and rolling friction test were carried out by high-speed photography technology, and the basic discrete element model contact parameters such as impact recovery coefficient, static friction factor, and rolling friction factor of banana straw were determined. Physical and simulated shear tests of banana straw were carried out to obtain the mechanical characteristic curve of destroying banana straw outer skin, and the maximum physical shear force was determined to be 122.41N. The optimal parameter combinations of normal contact stiffness, tangential contact stiffness, critical normal stress, and critical tangential stress of the banana straw bonding model were determined by central composite design (CCD) response surface method. The simulation results showed that the relative error between the simulated shear force results and the physical shear force was only 2.34%, which verified the reliability and accuracy of the bond parameter calibration method, and provided a theoretical reference for the design and research of the banana straw crushing and return machine.

Abstract:In order to solve the problem of low collection success rate caused by the attachment of tea leaves to the wall during selective picking of famous tea, a tea harvester was designed according to the physical parameters of one bud and two leaves. The factors that affecting the success rate of tea collection were obtained through pretest: the negative pressure value, the number of thread lines and the parameter variation range of thread lead. Fluent simulation and Box-Behnken response surface analysis were used to study the interactive influence of various factors on the success rate of tea collection. The experimental results showed that three factors changed the tangential velocity of air near the wall in the pipe;in the order of significance, the influence of each factor on the success rate of collection was negative pressure value, thread lead and thread number;taking the success rate of tea collection as the optimization goal, the parameters were optimized and rounded, and the parameters were obtained as follows: negative pressure value H=120Pa, thread number N=9, thread lead S=95mm. The optimized parameters were used to carry out adsorption collection test, the results showed that the success rate of tea collection under this condition was 98%, that was, the success rate of tea collection after optimization was 26 percentage points higher than that before optimization, and the relative error between the experimental value and the predicted value was less than 5%, so the optimization model was reliable.

Abstract:Aiming at the lack of entrainment loss detection technology in the current domestic corn kernel direct harvester kernel loss detection system, a corn kernel direct harvester entrainment loss detection system based on an embedded single-chip microcomputer was designed. The detection system included a loss detection sensor, a data collector and a data display terminal, which can monitor the cleaning loss and entrainment loss of the harvester at the same time, and feedback the harvesting loss rate and loss amount of the harvester in real time. The finite element analysis of monitoring boards of different materials and thicknesses was carried out through modal simulation software, and a 0.5mm thick 304 stainless steel plate was selected as the monitoring board;domestic EDA software was used to design signal processing circuits, including voltage followers and Butterworth filters, Multisim was used to simulate and analyze the performance of circuit;an adaptive time-limited filtering algorithm was designed based on the STM32 series single-chip microcomputer, which can effectively suppress the aftershock interference caused by the impact of grains and improve the accuracy of the sensor. On the test bench, under different conditions, the calibration test was carried out on corn kernels, residues and corn ears of different sizes to obtain signal characteristics;the installation test showed that the maximum error of the detection result of corn kernel entrainment loss was 9.96%, and the average error was about 6.52%, and the loss rate change trend feedback was timely, which can assist the staff in making operation decisions.

Abstract:The combine harvester knowledge base system uses the SQL server database, numerous data tables in the database are independent and easy to build and manage. But when the amount of knowledge base data reaches a certain size, querying data tables one by one is not actionable and merging all the data tables will lead to confusion in the data structure, unclear content expression, and technical inability to achieve. In response to this problem, a multi-table joint query method of combine harvester knowledge base data was proposed. The data table types was divided from multiple perspectives, the data storage structure of the combine harvester knowledge base was analyzed and the management scope for multi-table joint data was set. The application structured query language (SQL) fused multi-table information into a dataset and stored it into a temporary table to achieve multi-table joint operation. The human-computer interactive interface was used to convert the user query requirements into multi-table joint query statements to generate query results, and multi-table approximate range query and multi-table precise positioning query were realized. The test results between multi-table joint query and traditional single-table knowledge query showed that on the one hand, multi-table approximate range query saved user operation time by 50% or more than the original single-table approximate range query of the system, and the highest reached 90.4%;on the other hand, the multi-table precise positioning query saved 48.1% or more user operation time compared with the original single-table precise positioning query of the system, and the highest reached 89.6%. The implementation of multi-table joint query made the combine harvester knowledge base system practical and feasible and provided a reference idea and method for data management of similar knowledge base system.

Abstract:High-standard farmland construction is an important guarantee for national food security, and the quality assessment of high-standard farmland construction is beneficial to the implementation of farmland planning and government decision-making. As an important project of high-standard farmland construction, the rapid and accurate acquisition of field roads can provide basic data support for the quality assessment and effect evaluation of high-standard farmland construction. Thus, it is necessary to obtain accurate and effective field roads information. However, compared with high-grade roads, the narrow pavement width and easy occlusion by vegetation are the typical characteristics of field roads, which are the main factors leading to the low degree of automation in existing methods. Aiming at the problems of low accuracy and weak generalization ability of traditional recognition methods for narrow field roads, a highstandard farmland road recognition method was proposed based on U-Net network. Firstly, on the basis of analyzing the basic characteristics of the field roads, the GF-2 images were selected as the experimental data, and the object-oriented method was used to segment the image and classify it according to the characteristics of the object, so as to eliminate non-roads such as buildings with similar spectra elements to reduce interference;then, operations such as cropping, labeling, and data enhancement were performed on the image, the U-Net network was used to mine the deep and shallow features of the image, and the network was continuously trained by adjusting parameters to achieve accurate identification of field roads;finally, according to the characteristics of road breakpoints, the local connection method was used to repair the road breakpoints, and the accuracy verification were carried out in Dongting Town, Dingzhou City, Hebei Province as the experimental area. The results showed that by mining the image features of 622 field road samples, the U-Net network could effectively identify high-standard farmland roads in various scenarios. After repairing the road breakpoints, the field road identification precision in the study area reached 96%, and the recall and F1 score were 62% and 75%, respectively. The recognition accuracy could meet the requirements for rapid evaluation of high-standard farmland construction quality. Compared with traditional identification methods, the combination of object-oriented and deep learning methods could quickly identify field roads on the basis of reducing building interference, and could better solve the noise and misidentification issues caused by large differences in field road materials and vegetation occlusion. This method could provide a method reference for the identification of narrow objects in farmland.

Abstract:Human activities and climate affect land use change, and land use change is one of the most fundamental factors influencing habitat quality change. It is important to investigate habitat quality under different climate scenarios for sustainable use of regional land resources and ecological conservation. Taking Nanchang City as an example, and the shared socioeconomic pathways (SSPs) and representative concentration pathways (RCPs) were predicted based on coupled system dynamics (SD)-patch-generating land use simulation (PLUS) model simulations. The integrated valuation of ecosystem services and trade-offs (InVEST) model evaluated the land use pattern of Nanchang City in 2035 under combination of SSPs and RCPs. The results showed that under the three scenarios, the area of arable land, forest land and grassland in Nanchang City was decreased in 2035, the land for construction was expanded rapidly, and the change in water and unused land was small. Under the three climate scenarios, the habitat quality of Nanchang City in 2035 showed a decelerating trend of decline, mainly showing a shift from medium to low habitats, and the degradation degree from large to small was SSP585, SSP245, and SSP119. The research results can provide scientific reference for high-quality development and biodiversity conservation in Nanchang City, and it can also provide scientific reference for high-quality development and biodiversity conservation in Nanchang City.

Abstract:Maize leaf area index (LAI) displays a significant vertical distribution gradient. However, there is currently a limited amount of research focused on directly estimating the vertical distribution patterns of maize LAI from images. Designing an appropriate unmanned aerial vehicle (UAV) detection scheme can contribute to improving the accuracy of maize LAI estimation. Thus different maize varieties, and density and disease were used, and sowing experiments were carried out in the field to collect data on the vertical distribution of maize LAI. UAVs equipped with RGB, multi-spectral (MS), and thermal infrared (TIR) cameras were used to capture visible, multi-spectral, and thermal infrared images. Seven sets of UAV image data were collected during the reproductive growth stage of maize. To validate the effects of different UAV flight altitudes and solar zenith angles on maize LAI estimation, two completely controlled experiments with different flight altitudes were conducted, resulting in a total of 10 sets of UAV image data. Additionally, UAV image data were collected at each hour from 08:00 to 18:00 on a single day, resulting in 11 sets of image data, to discuss the robustness of the maize LAI estimation model under different flight experiments. A multi-source remote sensing image dataset was constructed to provide image feature variables highly correlated with maize LAI. Eight texture information categories were generated based on gray-level co-occurrence matrix from the original image texture features. In the end, 51, 43, and 9 image features were obtained from RGB, MS, and TIR image data sources, respectively. Seven machine learning models, including GBDT, LightGBM, MLPR, PLSR, RFR, SVR, and XGBoost, were selected to estimate the vertical distribution of maize LAI. These models were applied to estimate LAI vertical distribution data at different maize growth stages. Two models with the strongest robustness were selected to verify the optimal observation time and flight altitude under different drone flight heights and sun elevation angles. The research results showed that during the reproductive growth stage of maize, the best single growth period for estimating maize LAI was the tasseling period. The MLPR model had R2 of 0.91 and rRMSE of 5.1% for LAI estimation. At the same time, the LAI estimation accuracy obtained during the maize maturation period was the worst, with R2 of 0.8 and rRMSE of 11.01%. As the measurement height of maize LAI was increased, the accuracy trend differred from that at the bottom, showing a trend of first decreasing and then increasing. Based on the experiments conducted involving different flight and solar altitude angles, it was concluded that lower flight altitudes of UAVs led to higher accuracy in estimating maize LAI. Specifically, at a flight altitude of 30m, the MLPR model achieved an accuracy of R2 of 0.73 and RMSE of 10.97%. Additionally, the highest accuracy in maize LAI observation was achieved when observations were conducted at 09:00 and 10:00 in the morning. The use of UAV remote sensing technology, combined with multi-source image data, enabled accurate observation of the vertical distribution of LAI in maize canopies. This approach enabled a precise understanding of the spatial distribution of maize LAI at different heights, and provided timely information on the health status of functional leaves. The acquired data can be used to adjust field management measures accordingly. Furthermore, experts in maize breeding can use this technology to identify differences between maize varieties and select specific cultivars, which had significant practical implications.

Abstract:The physiological and biochemical properties of ramie are the result of comprehensive influence of genetic basis and environmental conditions, which can reflect ramie growth under specific stress environment. Therefore, a fast, accurate and inexpensive method is needed to monitor the dynamic changes of ramie physicochemical property during the whole growth cycle. Unmanned aerial vehicle (UAV) remote sensing technology provides an effective means for monitoring crop growth in large field, which has been widely concerned and applied by virtue of its advantages of fast, non-destructive, timely and accurate. However, at present, there are few researches on the comprehensive evaluation of ramie physicochemical property by using UAV multi-spectral images. The UAV was equipped with a multi-spectral camera to acquire the multi-temporal canopy images of ramie. Then, the canopy orthophoto image was obtained by DJI terra, and the spectral and texture characteristic values of ramie plants were further extracted. Pearson correlation analysis (PCA) and recursive feature elimination (RFE) were used to screen the sensitive eigenvalues. Finally, based on multi-temporal remote sensing data, linear regression (LR), random forest regression (RF), support vector machines (SVM), partial least squares regression analysis (PLSR) and decision tree (DT) were used to estimate ramie physicochemical property, respectively. The results showed that there was a significant correlation between the ramie physicochemical property and spectral skewness parameters. Both PCA and RFE can improve the accuracy of the estimation model, but RFE had better performance. The accuracy of the LR-SAPD estimation model was 0.662. The R2 and RMSE of LR-RWC estimation model were 0.793 and 2.213%, respectively. The SVR-LAI model could better estimate ramie LAI (R2=0.737, RMSE was 0.630). In conclusion, an accurate, efficient, cost-effective and universal dynamic monitoring method for physicochemical property of field ramie was proposed.

Abstract:The use of multispectral technology to carry out chlorophyll relative content (SPAD) monitoring of Cinnamomum camphora dwarf forest could provide timely diagnosis of Cinnamomum camphora dwarf forest growth and provide timely information support for field management decisions. The SPAD inversion model of Cinnamomum camphora dwarf forest was constructed by using UAV multispectral remote sensing images to extract band reflectance and filter vegetation index, which took band reflectance and vegetation index as model input respectively, and four methods were used: partial least squares regression (PLSR), support vector regression (SVR), back propagation (BP) neural network and radial basis function (RBF) neural network, and different input quantities and the inversion accuracy of simulation results of different models were compared. The results showed that there was little difference in the accuracy of inversion in the same model compared with two different inputs. Based on the partial least squares regression method, the estimation of SPAD of Cinnamomum camphora dwarf forest with vegetation index as the model independent variable was slightly better. Based on support vector regression, BP neural network and RBF neural network, the estimation of SPAD of Cinnamomum camphora dwarf forest with band reflectance as the model independent variable was slightly better. Compared with partial least squares regression, support vector regression and BP neural network, the accuracy of Cinnamomum camphora SPAD inversion based on RBF neural network was the highest. Taking the band reflectance and vegetation index as the input of the model as examples, the coefficient of determination (R2) was respectively 0.788 and 0.751, and root mean square error (RMSE) was respectively 1.838 and 2.457, indicating that RBF neural network had obvious advantages in predicting the SPAD of Cinnamomum camphora dwarf forest.

Abstract:Accurate extraction of ridges is an important prerequisite for digital agricultural management. However, due to the interference of factors such as occlusion and alopecia areata, it brings challenges for the semantic segmentation method to extract the ridge area. A U-Net segmentation network model was proposed based on a multi-information attention mechanism and an edge-aware module. Firstly, multi-information attention was introduced into the down-sampling of the U-shaped network to enhance the context information between adjacent layers and improve the representation ability of the semantic features of the ridge area. Secondly, the edge-aware segmentation module was applied to each layer of the U-Net decoding part, and the ridge edge information was extracted in different semantic feature layers to improve the semantic segmentation accuracy of the ridge region. Finally, the joint edge-aware loss and semantic segmentation loss were used to construct a joint loss function for overall network optimization. The training and model validation were carried out with the UAV wheat field data set collected by the wheat experimental base in Suixi County, Huaibei City, Anhui Province. The experimental results showed that the pixel accuracy of semantic segmentation of crop plants in different datasets was as high as 95.57%, and the average intersection ratio was 77.48%.

Abstract:Weight adjustment of sampling sites is a key aspect for spatial allocation of samples in the accuracy evaluation of remote sensing classification. Taking accuracy evaluation of sampling sites in Shunyi District of Beijing as an example, a weight adjustment method of sampling sites integrating area attribute and uncertainty information was proposed and named fuzzy adjustment weight method, which was used for sampling sites layout of accuracy assessment. Firstly, the fuzzy neutral index and its weight were constructed to stand for uncertainty information, and the fuzzy adjustment weight was constructed by fusing the fuzzy neutral index weight and area weight, and the fuzzy adjustment weight results of each stratum were calculated to achieve spatial allocation of samples in the feature space. Secondly, different gradient sample sets were drawn, and spatial-simulated annealing and the minimization of the mean of the shortest distances criterion were used to optimize sampling sites in geographical space. Finally, the indexes of weight adjustment evaluation were constructed to assess the effect of fuzzy adjustment weights. The comparative analysis was achieved between fuzzy adjustment weight and the other weight adjustment methods and the methods without weight adjustment. The results showed that the fuzzy adjustment weights of large, medium and small uncertainty strata in Shunyi District were 0.45, 0.37 and 0.18, respectively. Compared with the area weights of each stratum, the weights of large and medium uncertainty strata were increased significantly and slightly, respectively, and the weight of small uncertainty stratum was decreased significantly. The overall accuracy, relative accuracy, root mean square error and standard deviation of the accuracy evaluation results for weight adjustment of five different sample sets were 69.90%~73.48%, 96.28%~99.82%, 0.01 and 0.01, respectively. The evaluated effect of fuzzy weight adjustment method was better than the methods with area weight, fuzzy neutral index weight, uncertainty stratification weight, and the spatial even sampling and simple random sampling methods. The weight adjustment of sampling sites for the developed method was more accurate and reliable. The developed fuzzy adjustment weight method used for sampling sites layout of accuracy assessment can integrate the area attribute and uncertainty information, and avoid excessive weight adjustment, which was used to improve the rationality for spatial allocation of sampling sites in each stratum.

Abstract:In order to accurately detect litchi diseases and insect pests with complex background in natural environment in real time, the data set of litchi diseases and insect pests was constructed and the detection model of litchi diseases and insect pests was proposed for diagnosis and control. Based on YOLO v4, GhostNet, the lighter and faster lightweight network, was used as the backbone network to extract features. According to the core design of GhostNet, Ghost Module, a lower cost convolution, was used to replace the traditional convolution in the neck structure. Based on the lightweight YOLO v4-G model, the feature fusion method and attention mechanism called CBAM were used to improve the YOLO v4-G. The detection accuracy was improved without losing the detection speed and the lightweight degree of the model. Finally, the YOLO v4-GCF detection model of litchi diseases and insect pests was proposed. The dataset contained 3725 images of litchi diseases and insect pests. Litchi diseases included sooty mold, anthracnose and algal spot. Litchi insect pests included leaf mite and Dasineura sp. The experimental results showed that the average accuracy of five kinds of diseases and insect pests targets detected by YOLO v4-GCF detection model in train set, validation set and test set was 95.31%, 90.42% and 89.76%, respectively. The detection time of a single image was 0.1671s, and the size of the model was 39.574MB. Compared with the YOLO v4, the model size was reduced by 84%, the detection speed was increased by 38% and the average accuracy in the test set was improved by 4.13 percentage points. At the same time, the average accuracy was 17.67,12.78 and 25.94 percentage points higher than those of YOLO v4-tiny, EfficientDet-d2 and Faster R-CNN, respectively. The proposed YOLO v4-GCF detection model of litchi diseases and insect pests can effectively inhibit the interference of complex background, and accurately and quickly detect targets of litchi diseases and insect pests in the images, which can provide reference for crop diseases and insect pests detection research with complex and unstructured background in natural environment.

Abstract:In order to realize the rapid and accurate recognition of greenhouse tomato fruit by agricultural picking robot in the complicated environment of greenhouse, a fast target detection method for greenhouse tomato fruit based on attention mechanism and improved YOLO v5s was proposed. According to the characteristics of small size and fast speed of YOLO v5s(You only look once v5s) model, the convolutional block attention module (CBAM) was added into the backbone network. By concatenating spatial attention module and channel attention module, the problem of color similarity between green tomato fruit and its background was solved. More attention was paid to the target features of green tomato fruit to improve the recognition accuracy. Replacing GIoU Loss with CIoU Loss as the new loss function of the algorithm contributed to improve the positioning accuracy while improving the bounding box regression rate. The test results showed that the recognition accuracy of the CB-YOLO network model for red tomato fruit detecting precision and green tomato fruit detecting precision and mean average precision in greenhouse environment was 99.88%, 98.18% and 99.53%, respectively. Compared with Faster R-CNN network model, YOLO v4-tiny network model and YOLO v5 network model, the detection accuracy and the mean average precision were improved. The CB-YOLO model was deployed to Android system of mobile phones after being tested by different mobile phones, which verified the stability of the performance detection of the deployment model under actual working condition. It will provide technical support for target detection and harvesting based on robotic mobile edge computing in facility environments.

Abstract:With the continuous development of smart agricultural technology, plant disease identification models are increasingly pursuing the goals of accuracy, efficiency, and light weight. Aiming at the problems of large number of parameters, high calculation cost and low accuracy in the current tomato disease recognition model, a lightweight network was proposed based on multi-scale feature fusion and coordinate attention mechanism (Multi-scale feature fusion and coordinate attention MobileNet, MCA-MobileNet) model. Totally ten types of tomato leaf images were collected, and Wasserstein generative adversarial networks (WGAN) based on Wasserstein distance for data enhancement was used, which solved the problem of insufficient and unbalanced sample data and improved the generalization ability of the model. On the basis of the original model MobileNet-V2, an improved multi-scale feature fusion module was introduced to extract features from feature maps of different scales to improve the adaptability of the model to different scales;the lightweight coordinate attention mechanism module (Coordinate attention, CA) embedded in the inverted residual structure, so that the model paid more attention to the disease characteristics in the leaves and improved the recognition accuracy of the disease types. The test results showed that the accuracy rate of MCA-MobileNet for identifying tomato leaf diseases reached 94.11%, which was 2.84 percentage points higher than that of the original model, and the number of parameters was only 1/6 of the original model. This method better balanced the recognition accuracy and calculation cost of the model, and provided ideas and technical support for field deployment and real-time detection of tomato leaf diseases.

Abstract:The flowering intensity of litchi can directly affect the yield and quality of the fruit, so the detection of litchi flowers is very important for orchard planting strategies. Dense litchi flower detection has important challenges due to serious occlusion. Existing research methods ignore the interaction between dense suggestion boxes in the process of non-maximum suppression. In order to improve the detection precision and reduce the missed detection rate, a detection method was proposed based on polyphyletic loss function. This method included an aggregation loss term in the loss function to force the proposal box to approach and compactly locate the corresponding object. At the same time, the segmentation loss of the bounding box specially designed for the dense crop scene was added to keep the prediction box away from the surrounding objects and improve the robustness of detecting a large number of flowers. Compared with Faster R-CNN, Focus Loss, AdaptiveNMS and Mask R-CNN, the experiment showed that the recognition precision of this method on the standard apple blossom dataset was about 2 percentage points higher than that of other methods, which verified the detection versatility of this method. At the same time, the mean average precision of this method in the self-built litchi flower dataset was 87.94%, the F1 score was 87.07%, and the miss rate was 13.29%. Compared with Faster R-CNN, Focus Loss, AdaptiveNMS and Mask R-CNN, the accuracy of the method was improved by 20.09 percentage points, 14.10 percentage points, 8.35 percentage points and 4.86 percentage points, respectively, with high detection performance. Therefore, the method proposed can effectively detect the dense litchi flowers, and provide an important reference for dense crop detection in complex scenes.

Abstract:Accurate and efficient cow behavior recognition is helpful for timely disease detection and detection of abnormalities. It is the key to perceive cow health. By analyzing the behavior of cows at different periods in the cattle farm, a cow behavior recognition algorithm based on spatiotemporal features was proposed. The algorithm combined temporal shift module (TSM), feature attention unit (FAU) and long short-term memory (LSTM) networks on the basis of time-domain segment network (TSN). Firstly, TSM was used to fuse time information to improve timing modeling ability. The video frame after time sequence modeling was input to TSN. Secondly, FAU was used to integrate high resolution spatial information and low resolution semantic information to enhance the learning ability of spatial features of the algorithm. Finally, the past and current information were fused by LSTM to classify cow behavior. The results showed that the recognition accuracy of this algorithm for eating, walking, lying, and standing was 76.7%, 90.0%, 68.0% and 96.0%, respectively. And the average recognition accuracy was 82.6%. Compared with C3D, I3D and CNN-LSTM networks, the average recognition accuracy of this algorithm was 7.9 percentage points, 9.2 percentage points and 9.6 percentage points higher, respectively. The illumination variation had a certain impact on the recognition accuracy, but the proposed algorithm was relatively less affected by light. The results can provide technical support for cow health perception and disease prevention.

Abstract:To address the current problem of low accuracy in the recognition of feather pecking anomalies (including pecking and pecked) in laying hens, a method for feather pecking anomaly recognition was proposed based on an improved YOLO v6-tiny model. By introducing the DenseBlock structure into the YOLO v6-tiny model and incorporating the SPP module SPPCSPC into the CSP structure, the feature extraction capability of the YOLO v6-tiny model was enhanced, the sensory field of the model was expanded, and the detection accuracy of the model was improved. Based on the identification of feather pecking anomalies, how to classify individual laying hens was investigated based on the number of anomalies. The method to identify individual laying hens based on the YOLO v6-tiny model was proposed and identification results of feather pecking anomalies were input into the individual identification network to classify individual laying hens. At the same time, the change pattern of the number of anomalies at two different breeding densities and three different times of the day was also analyzed. The experimental results showed that the average precision (AP) of the optimized model were 92.86% and 92.93% for pecking and pecked anomalies, respectively, which were 1.61 percentage points and 1.08 percentage points higher than that of the YOLO v6-tiny model, 3.28 percentage points and 4.00 percentage points higher than that of the Faster R-CNN model, 6.15 percentage points and 6.63 percentage points higher than that of the YOLO v4-tiny model, 2.04 percentage points and 4.27 percentage points higher than that of the YOLO v5s model, and 5.39 percentage points and 3.92 percentage points higher than that of the YOLO v7-tiny model. The method can identify the abnormalities of pecking and pecked feathers, which provided technical support for the intelligent detection of abnormal behavior of laying hens. The results of classifying individual laying hens based on pecking abnormalities provided a basis for preferential breeding of individual laying hens.

Abstract:To recognize an individual goat under farm conditions, a novel goat face recognition model named DWT-GoatNet was proposed based on wavelet transform and convolutional neural networks, which integrated frequency domain features and spatial domain features. Firstly, facial images of a total of 30 highly similar Xinong Saanen dairy goats were collected under two different light conditions, daytime and night. Some images were discarded based on structural similarity (SSIM), and the remaining images were cropped manually. Image sets were also augmented by operations of blur, brightness adjustment, translation, rotation, noise addition and scaling. Secondly, a goat face feature extraction module was designed based on twodimensional discrete wavelet transform (2D-DWT) and convolution operation to achieve feature fusion. Then, with this module, a classification module was added and a convolutional neural network named DWT-GoatNet was built. Finally, the combination of hyper-parameters was optimized and goat face recognition model was formed. The experimental results showed that the accuracy of the proposed goat face recognition model can reach 99.74% and 99.89%, respectively, on test set under different light conditions of daytime and night, which was higher than that of some classical CNNs such as AlexNet, VGGNet-16, GoogLeNet, ResNet-50 and DenseNet-121, while the DWT-GoatNet can provide an effective recognition for some related fields of precision farming and agricultural insurances.

Abstract:In order to meet the demand of sustainable precision agriculture for efficient and realtime acquisition of agricultural production information, flexible pH sensors have been widely studied for their small structure, easy integration and strong mechanical performance. However, the problem of deviation between multi-layer printing in the preparation process has not been solved. Therefore, aiming at improving the accuracy of sensor preparation, the registration performance of affine transformation model was discussed based on MSER. The model verification and precision test experiments were carried out to verify the feasibility of the registration model and the accuracy that the registration can achieve. The model was applied to the preparation of the flexible pH sensor chip, and finally the feasibility of the registered flexible pH sensor chip in the online monitoring of tomato organic matrix pH was verified. The experimental results showed that the accuracy of the model can reach the minimum line width of 90μm. Above 90μm, the RMSE and MAE were not more than 140μm, MRE was 3.33%~22.22%. The minimum line spacing was 500μm. Above 500μm, the RMSE and MAE were not more than 160μm, MRE was 11.67%~24%. The sensitivity of the flexible pH sensor prepared by this registration method was -61.9mV per unit pH value, and the response range was 2.0~10.0. Compared with the standard glass pH sensor, the absolute error was less than 0.15 and the relative error was less than 4.1%. The slope of the fitting curve was 0.97 and the determination coefficient R2 was 0.99. The fitting degree and accuracy was high. In the pH monitoring test of tomato organic matrix, the measurement results of the flexible and standard pH sensors had good consistency. The absolute error of the synchronous measurement results was less than 0.09, the relative error was less than 1.5%, and the RMSE between the pH monitoring results for a continuous week was only 0.05. The registered flexible pH sensor chip showed a good agricultural application prospect because of its miniaturization and close performance to the standard pH sensor. This registration method had a certain reference value for the preparation of flexible sensors processed by multiple processes, and provided a strong preparation for the subsequent monitoring of flexible pH sensor chips in the crop growth environment.

Abstract:In the context of global change, it is important for sustainable agricultural development to reveal the response of soil nutrient system to changes in climate and human cropland use. Based on the climate, land use intensity (LUI) and soil nutrient data of Yanqi Basin, the sensitivity, contribution rate and GIS methods were used to analyze the sensitivity and spatial differentiation of soil nutrient system to climate and human land use activities. The results showed that there was an obvious trend of warming and wetting in Yanqi Basin (the temperature tendency rate was 3℃/(10a), the precipitation tendency rate was 39mm/(10a)), and the LUI was increased slightly. There were significant spatial differences in the sensitivity and response of soil nutrient systems to climate and LUI changes. The soil nutrient system had higher spatial heterogeneity in sensitivity to temperature change and more depth of response to LUI change. The contribution rate of climate and LUI to soil nutrient system was significantly different, and the contribution rate of LUI was greater than that of climate. The sensitivity of soil nutrient system and its spatial distribution were the result of mutual regulation and spatial differentiation between natural environment and human cultivated land utilization. The results can provide scientific basis for improving the ability of soil nutrient system to cope with climate change and the precise and efficient management of cultivated land.

Abstract:The undisturbed soil samples from the profiles of different reclamation areas in Rudong County, Dongtai City and Binhai County of Jiangsu Province were taken as the research objects. Based on CT scanning technology, through the construction of pore network model, the soil pore distribution and its topological relationship were quantitatively characterized, and the soil pore throat parameters, including pore size distribution, coordination number, throat radius and throat length, were statistically analyzed. Using the extracted pore network, the seepage simulation of single-phase water in soil was realized, the absolute permeability coefficient (saturated hydraulic conductivity) of soil and its main influencing factors were analyzed. The results showed that the reclamation activities played a positive role in improving soil pore activities. The specific performance was that with the increase of reclamation years, the radius range, average coordination number and throat radius of connected pores tended to increase, while the throat length tended to decrease. The pore structure of shallow soil layer was better. In single-phase water simulation, the simulated permeability coefficient was in good agreement with the measured saturated hydraulic conductivity, showing a consistent change trend. The grey correlation degree showed that the throat radius had the greatest influence on soil permeability.

Abstract:A trapezoidal fuzzy numbers and distributed crop simulation model was developed based on distributed agricultural production warning model. The model selected spatially-distributed crop yield and water productivity as warning indictors to calculate system’s warning levels. Besides, it introduced trapezoidal fuzzy numbers to express uncertainties of crop yield target and water productivity target. Moreover, the crop yield and water productivity between 1976 to 2021 and between future year 2026 to 2045 under RCP4.5 scenario under normal irrigation, 0.8 times and 0.6 times as normal irrigation were simulated based on distributed AquaCrop model. Additionally, the static and dynamic coordinated degrees between crop yield and water productivity for future 20 years were calculated. The results showed that warning level of one crop with different soil types and irrigation levels was different. And warning levels presented irregular changes with decrease of irrigation level while coordinated level lessened with decrease of irrigation level. The model could identify warning levels of spatially-distributed crop yield and water productivity under different irrigation levels and realize the precise warning forecast, which were beneficial to reach reductions of agricultural production risks quickly.

Abstract:Soil hydrodynamic parameters are the basic parameters for simulating the process of soil material transport in the field. Accurate determination of soil hydrodynamic parameters is of great significance to achieve precise regulation of farmland habitat. For one-dimensional vertical infiltration experimental data，based on algebraic and numerical methods, three different objective functions were constructed, and the applicability of the whale optimization algorithm and grey wolf optimizer was analyzed to invert the parameters of the Brooks-Corey-Mualem model. The result showed that by choosing an appropriate objective function, both swarm intelligence optimization algorithms can be used to invert soil hydrodynamic parameters. In the algebraic method, the whale optimization algorithm optimized the soil hydrodynamic parameters with the fixed parameters θr and θs under the objective function two (relative error composed of cumulative infiltration, time, and soil water content profiles) with the smallest error. The relative errors of the cumulative infiltration volume, infiltration rate, and soil water content profiles obtained from the inversion parameters were all below 9.74%, the determination coefficients were all above 0.9040, and the inversion time was 70s. In the numerical method, the parameter error derived from the fixed parameters θr and θs under the objective function three (relative error composed of cumulative infiltration, depth of wetting front, and soil water content profile) of the grey wolf optimizer was the smallest. The relative errors of the cumulative infiltration volume, infiltration rate, and soil water content profiles obtained from the inversion parameters were all below 2.53%, the determination coefficients were all above 0.9917, and the inversion time was 115s. Therefore, the algebraic method took a short time and has relatively low accuracy, while the numerical method took a long time and has a relatively high accuracy. When inverting soil hydrodynamic parameters, an appropriate optimization method can be selected according to the error accuracy requirements.

Abstract:The cracks caused by drying and shrinkage in farmland affected the soil hydraulic and physical structure characteristics and provided preferential channels for the transport of irrigation water or pollutants. In order to reveal the evolution characteristics of crack morphology and shrinkage of matrix in farmland and predict the crack porosity, experiments were carried out in a greenhouse. Digital image processing and morphological algorithm were used to analyze the geometric characteristics of cracks during the dry-wet cycle.Non-invasive local shrinkage analysis method was used to quantify the shrinkage characteristics of matrix domain. According to the transformation mechanism of soil pores from matrix domain to subsidence domain and crack domain during dewetting process, a prediction model of soil crack porosity was proposed, which included two submodels: VG-PENG model and Logistic function of shrinkage geometric factor. Based on the measured crack porosity experiment, the universal applicability of the model was verified for the crack porosity of soil samples of different thicknesses. From the perspective of soil physics, the evolution process of crack porosity with respect to water content was effectively predicted. The results showed that layer thickness had significant influence on crack morphology during soil cracking. With the increase of soil layer thickness, the width of large cracks on the surface of the reconstituted soil was increased, the possibility of secondary cracking was decreased, and the density of cracks was decreased, and the process of soil shrinkage and cracking was slowed down. In the process of drying and shrinking, the soil showed the phenomenon of agglomeration to block core. The minimum value deformation near the core in the shrinkage block area tended to be 0, and the edge wall area of cracks showed a concentrated deformation area, and the shrinkage was increased with the increase of soil layer thickness. The vertical uniformity of drying rate and drying degree also changed with soil layer thickness. Thicker soils usually had wider cracks in the soil surface and larger clumps of soil formed by cracking. In the cracking process, the shrinkage anisotropy showed vertical subsidence in the first stage, vertical subsidence in the middle stage with slight horizontal shrinkage, and horizontal cracking in the late stage. Considering the anisotropy of soil shrinkage, the prediction model of crack porosity proposed made up for the shortcoming that the previous crack models did not consider the shrinkage anisotropy, and the model had good effect and was generally applicable in different soil thicknesses (R2>0.91). The study quantified the development characteristics of crack and matrix at soil samples of different thicknesses, which provided the parameter basis for the construction of crack flow model and the hydraulic characteristics of soil matrix, and contributed to the formulation of expanded soil remediation and water management schemes.

Abstract:The front roof is an important part of realizing the communication between the internal and external environment of the Chinese solar greenhouse (CSG). The internal environment of CSGs is controlled by the state of the front roof covers in actual production. Three important covers (insulation quilt, transparent film and colored mesh) are laid on the arc-shaped front roof to control the internal environment factor of CSGs, such as the light, temperature and humidity. The rolling and laying of three covers usually rely on the common front roof structure surface. Due to the coincidence of action trajectory range and operation methods, there are problems that interfere with each other and lead to the low degree of automation in the actual production. It is also not conducive to the systematic and intelligent control of CSGs. According to the functional requirements of the three covers, a coordinative operation device was designed. The main research content included two hardware parts (the outer operation structure and the inner operation structure) and their coordinated control method. The specific corresponding transmission structure and the position detection method of covers were emphasized. The key points of this research were verified by experiments of the scaled model and field CSG. The experiments result of the scaled model showed that the mean relative error (MRE) of different opening values of three covers was less than 3.5%, the root mean square error (RMSE) was less than 2.0%. The value of coefficient of variation (CV) of the device was less than 3.0% at each opening point. The outer operation structure also showed good control accuracy and stability in the field experiments, the MRE was less than 2.5%, RMSE was less than 1.26%, and the CV was less than 1.0%. The overall measured experiment results of the scaled model and field CSG showed similar changing trends, which verified the model experiment results and the implementation feasibility of the overall designed device. The research provided a theoretical basis and systematic technical support for the operation of three covers, which was of great practical significance to the automation and intelligent development of the CSG.

Abstract:Differences in geographic location and environmental factors lead to differences in rice quality. Rice of high quality origin has excellent quality and taste and is more attractive to consumers, so it is important to study the origin traceability of rice and establish a credible traceability system of rice origin. The traditional IoT blockchain traceability system uploads traceability data to a centralized server, which in turn uploads it to the blockchain;this does not make good use of the resources in the edge nodes and also makes it vulnerable to security risks such as data forgery or data loss. A set of rice origin traceability model was designed based on blockchain and edge computing. The edge layer of this model divided the local blockchain network according to the origin latitude and longitude, each origin region chain contained multiple edge nodes and one core node, and relied on the edge computing capability of embedded devices for real-time data fusion of sensor data and deployment of blockchain on embedded devices. Edge computing ensured the authenticity of the data source, and the blockchain ensured that the on-chain data cannot be tampered, and the combination of the two realized the trusted storage of traceability system data. In addition, the storage scaling method of the blockchain network under the edge computing scenario and edge computing workflow were designed. Finally, after testing and analysis, the average time for querying public traceability data was 45.84ms, the average time for querying private traceability data was 50.92ms, the average time for encrypting onchain of edge nodes was 1.27s, and the storage capacity consumption of edge multi-chain was 18% of that of traditional single-chain, which can meet the practical application requirements of rice origin traceability.

Abstract:To meet the privacy protection problem of amount and identity in kiwifruit industry consortium blockchain transactions, a kiwifruit industry chain privacy transaction scheme based on +HomElG-ZKProof (+HomElG zero knowledge proof) and SM2 was proposed. Firstly, the transaction amount with +HomElG was encrypted and sent to the receiver by the transferor, and the signature based on the SM2 to confirm the transaction was generated and sent to the transferor by the receiver. Secondly, zero-knowledge proof evidences for the ciphertext related to the transaction amount based on +HomElG-ZKProof, ring signatures for the ciphertext related to the amount and the identity of the transaction based on the SM2 linkable ring signature were generated, with the receiver’s SM2 signature was packaged and uploaded to the consortium blockchain through the system layer Raft consensus by the transferor. Then the SM2 signature, the two ring signatures, and the link to confirm the transaction identity were verified by the supervisory node, the PBFT consensus at the application layer was used by verifying the ciphertext related to the transaction amount, ring signature and the zeroknowledge proof evidence related to the transaction amount to confirm the validity of the transaction by the user nodes. Finally, the valid transaction block number through the Raft consensus of the system layer was uploaded and the account balance was updated by the supervisory node. The analysis showed that the proposed scheme had the advantages of anti tamper attack, anti public key substitution attack, anti counterfeiting attack and anonymity, and had higher security. The test results showed that the scheme can realize double privacy protection of transaction amount and identity of users in the kiwifruit industry consortium blockchain. The experimental results showed that when the security parameter was 2048bit, the transaction time took about 4.495s, it can meet the actual needs of kiwifruit industry consortium blockchain transactions.

Abstract:The CIELAB color characteristics of 175 dry red wine samples from Northwest China were studied. Based on the visual characterization method of red wine color, the micro-quantitative classification and macro quantitative gradation of color quality of the tested wine samples were conducted. The results showed that the color characteristics of all wine samples had a certain degree of dispersion and differentiation, indicating that the characteristics of wine samples’ color were diversity, uniqueness and difference, which could be used as the basis of micro quantitative classification and macro quantitative gradation. The three color attributes (hue, chroma and lightness) were divided into five sub-classifications respectively according to their value ranges of 175 red wine samples. Totally 125 micro color classifications of wine samples were obtained, which could be applied to the digital description, transmission and comparison of color characteristics of red wines. Based on the sensory tendency of red wine color to build red wine color quality quantitative gradation method, wine samples were divided into six macro color grades. Compared with other classification and gradation methods, the micro-quantitative classification and macro-quantitative gradation method was simple, easy, standardized, objective and digital, and its grading effect was obvious, implying the value of application and promotion of the method was high.

Abstract:With the aim to investigate the effects of adding three polysaccharide types and concentrations of pectin, xanthan gum and κ-carrageenan (0.002g/mL，0.004g/mL，0.006g/mL) on the gel properties and molecular forces of soy protein (SPI)-chlorogenic acid (CA) hydrogels, the infrared spectra, molecular forces, rheology, water distribution and microstructure of SPI-CA/polysaccharide ternary gels were measured. Finally, the interaction mechanism between different polysaccharides and SPI-CA and the differences and similarities of different polysaccharides on the improvement of SPI-CA gel properties were confirmed. The results showed that the addition of xanthan gum and κ-carrageenan increased the β-sheets content of the protein from 24.59% to 24.87%~26.65% and significantly decreased the content of random coil (P<0.05), which was beneficial to the formation of the gel network structure. The addition of all three polysaccharides significantly enhanced the hydrogen bonding interactions in the gels and the binding ability of immobilized water (T21) in the gels. Compared with pectin and xanthan gum, the addition of κ-carrageenan resulted in stronger hydrogen bonding interactions in the gels, shorter relaxation time of T21, and denser gel structure. The storage modulus (G′), loss modulus (G″), hardness and chewiness of the gels were positively correlated with the concentration of κ-carrageenan. The addition of appropriate amounts of pectin and xanthan gum can form stable gels, but overfilling with pectin and xanthan gum can reduce the thermal stability of the gels and destroy the gel network structure. The results can provide theoretical support for introducing different polysaccharides to improve the gel properties of soy protein-chlorogenic acid hydrogels. 

Abstract:A parallel diesel-electric hybrid tractor was taken as the research object. By analyzing the topological characteristics and working characteristics of hybrid tractors, the coupling dynamic model of tractor rotary tillage unit and ploughing unit was constructed. With the goal of minimizing the equivalent fuel consumption of the whole machine, the motor power and diesel engine power were the control variables, and the state of charge（SOC）of the battery was the state variable, and the energy-saving control strategy based on Pontryagin’s minimum principle（PMP）was established. Finally, the Matlab simulation platform was used to simulate the energy-saving control strategy based on the Pontryagin’s minimum principle for the two typical working conditions of the tractor. The control strategy based on the optimal economic curve was used as a comparison strategy to verify the control effect based on the Pontryagin’s minimum principle. The analysis and comparison of the results of these two control strategies showed that: under rotary tillage conditions, the SOC value based on the optimal economic curve control strategy was decreased rapidly, while the SOC value based on the PMP control strategy was decreased slowly. Under the ploughing condition, the SOC value based on the optimal economic curve control strategy was both increased and decreased by a large margin, while the SOC value based on the PMP control strategy was only decreased. It was shown that the energy management strategy based on the Pontriagin minimum principle can reasonably control the operating state of diesel engine and motor, so that diesel engine and motor work in highefficiency areas. It effectively reduced the equivalent fuel consumption of tractors in the field. Under rotary tillage, the equivalent fuel consumption was reduced by 10.44%;under the ploughing condition, the equivalent fuel consumption was reduced by 11.20%.

Abstract:The kinematics and dynamics characteristics of a double-ring multi-rod antenna deployment mechanism were analyzed. Firstly, the configuration characteristics of the deployment mechanism were analyzed, and the unit decomposition of the whole mechanism was carried out. Based on the screw theory, the screw constraint diagram of the unit mechanism was drawn. Then the screw constraint equations of the unit mechanism were established, and the angular velocity and the linear velocity of the center of mass of the components in the mechanism were obtained by the screw velocity recursion. The velocity Jacobian matrix was obtained by the coordinate transformation derivative method. The screw acceleration and angular acceleration of the components and the linear acceleration of the center of mass were obtained by using the screw derivative and the screw acceleration synthesis rule. Finally, based on the Lagrange equation, the dynamic equation of the whole mechanism was established. Through numerical calculation and simulation, the kinematic and dynamic characteristics of different components were analyzed, and the correctness of the theoretical derivation was verified. The overall structure of the double-ring multi-rod antenna deployment mechanism analyzed was simple. It was a single-degree-of-freedom mechanism that can be folded and unfolded with only one drive. The kinematics and dynamics modeling and analysis method based on the combination of screw theory and Lagrange method had clear physical meaning, it can be better applied to the analysis of such space deployable mechanisms.

Abstract:In order to ensure the precise picking of tea picking robot, the 6-DOF hybrid tea picking robot mechanism installed on the mobile vehicle was studied and designed. According to the topological structure design theory of robot mechanism based on orientation feature, a 6-DOF hybrid manipulator was proposed and designed. The correctness of the mechanism was confirmed by analyzing the position of the forward and inverse solutions of the hybrid mechanism. Taking the minimum sum of lengths of links as objective function, the nonlinear programming method was used to optimize the dimensions of the links of the mechanism, and the optimal dimensions of all the links were obtained. The mass parameters and position parameters of the counterweight under complete balance were obtained by using the finite position method, so that the swing was constant to 0 and the position of the total center of mass of the mechanism was constant. The genetic algorithm was used to study the partial balance optimization of the swing force, and the optimal solution of its counterweight mass and position parameters was obtained, and the minimum total swing force was obtained within a certain range. The partial balance optimization results showed that the fluctuation of the centroid trajectory in the y and z-directions was decreased by 53.72% and 25.10%, respectively, and the total swing force fluctuation was decreased by 43.33%, which verified the effectiveness of the partial balance optimization. Based on the finite position method, the steps of action balance analysis laid a technical foundation for the structural design and prototype development of the hybrid tea picking robot. 