王海江,蒋天池,YUNGER John A,李亚莉,田甜,王金刚.基于支持向量机的土壤主要盐分离子高光谱反演模型[J].农业机械学报,2018,49(5):263-270.
WANG Haijiang,JIANG Tianchi,YUNGER John A,LI Yali,TIAN Tian,WANG Jin’gang.Hyperspectral Inverse Model for Soil Salt Ions Based on Support Vector Machine[J].Transactions of the Chinese Society for Agricultural Machinery,2018,49(5):263-270.
摘要点击次数: 2191
全文下载次数: 1042
基于支持向量机的土壤主要盐分离子高光谱反演模型   [下载全文]
Hyperspectral Inverse Model for Soil Salt Ions Based on Support Vector Machine   [Download Pdf][in English]
投稿时间:2018-01-31  
DOI:10.6041/j.issn.1000-1298.2018.05.031
中文关键词:  土壤  盐分离子  新疆  高光谱  反演模型
基金项目:国际科技合作项目(2015DFA11660)、石河子大学校级项目(RCZX201522)和石河子大学大学生研究训练计划项目(SRP2017024)
作者单位
王海江 石河子大学 
蒋天池 石河子大学 
YUNGER John A 州长州立大学 
李亚莉 新疆农业科学院 
田甜 石河子大学 
王金刚 石河子大学 
中文摘要:快速、无损、定量地获取土壤盐分离子组成及含量是盐渍化土壤治理、改良和利用的重要依据。以新疆盐渍化土壤为研究对象,应用高光谱分析技术获取不同区域土壤盐分离子的特征光谱,在对光谱数据去噪、数据变换基础上分析了鲜样(T1)、风干(T2)和干燥(T3)3类土壤,过2、1、0.15mm筛处理对离子含量光谱拟合模型精度的影响,建立了基于支持向量机的土壤主要盐分离子光谱反演模型,并对模型的精度和普适性进行了检验。结果表明:土壤原始特征光谱与盐分离子含量均不存在显著相关性,最大相关系数为Na+的0.41;通过光谱数据变换能够明显增强特征波段与离子含量的相关性,K+、Na+、Mg2+、Ca2+、SO2-4、Cl-和HCO-3的最优变换形式分别为(lgR)′、(lgR)′、R′、(lgR)′、CR、R′和CR,T1处理构建的拟合模型均不能很好地反演离子含量,T3处理的模型估测精度优于T2,土壤粒径越细对土壤离子含量的光谱反演效果越好。分析各处理模型的决定系数和标准误差表明,经T3处理、过0.15mm筛所构建的离子拟合模型预测精度最高,其中K+、Na+、Ca2+和SO2-4的RPD分别为2.153、2.6745、2.051和2.7864,以未参与建模和检验的石河子垦区土样对4种离子模型的普适性检验,其R2分别为0.6214、0.6897、0.6144和0.6507,说明构建的模型适于估算该区域土壤K+、Na+、Ca2+和SO2-4的含量。
WANG Haijiang  JIANG Tianchi  YUNGER John A  LI Yali  TIAN Tian  WANG Jin’gang
Shihezi University,Shihezi University,Governors State University,Xinjiang Academy of Agricultural Sciences,Shihezi University and Shihezi University
Key Words:soil  salt ions  Xinjiang  hyperspectral  inverse model
Abstract:The rapid, nondestructive and quantitative analysis of the composition and content of soil salt ions is an important basis for the treatment, improvement and utilization of salinized soil. Taking the saline soil of Xinjiang as the research object, the hyperspectral analysis technique was used to obtain the spectral characteristic of soil salt ions in different regions;compared the effect on the accuracy of the soil salt ions fitting model in different treatments which included fresh soil, air drying and oven dry sample and different particle sizes (2mm, 1mm and 0.15mm, respectively). After the transformation of spectral data, the spectral inversion models of main salt ions were established based on support vector machine (SVM), and the accuracy and universality of the model were tested. The results showed that there was no significant correlation between the original spectral characteristics and soil salt ions content, and the maximum correlation coefficient was Na+ (R=0.41). It was clear that the spectral data transformation can significantly enhance the correlation between the characteristic bands and the ions content, the optimal transformation forms of K+,Na+, Mg2+,Ca2+, SO2-4, Cl-and HCO-3 were (lgR)′, (lgR)′, R′, (lgR)′, CR, R′ and CR, respectively. The fitting models of T1 treatment cannot inverse the ions content very well, the accuracy of T3 model was better than that of T2, and the smaller the soil particle size was, the better the spectral inversion effect of soil ions content was. The prediction accuracy of the ions fitting model was the highest by T3 and over 0.15mm sieves, the relative prediction deviation of K+, Na+, Ca2+ and SO2-4 were 2.153, 2.6745, 2.051 and 2.7864, respectively,the universality test of four ion models was carried out by using samples from Shihezi area other than modeling and validation, the R2 of test models were 0.6214, 0.6897, 0.6144 and 0.6507, respectively. The models were suitable for estimating the content of soil K+, Na+, Ca2+ and SO2-4 in Xinjiang area.

Transactions of the Chinese Society for Agriculture Machinery (CSAM), in charged of China Association for Science and Technology (CAST), sponsored by CSAM and Chinese Academy of Agricultural Mechanization Science(CAAMS), started publication in 1957. It is the earliest interdisciplinary journal in Chinese which combines agricultural and engineering. It always closely grasps the development direction of agriculture engineering disciplines and the published papers represent the highest academic level of agriculture engineering in China. Currently, nearly 8,000 papers have been already published. There are around 3,000 papers contributed to the journal each year, but only around 600 of them will be accepted. Transactions of CSAM focuses on a wide range of agricultural machinery, irrigation, electronics, robotics, agro-products engineering, biological energy, agricultural structures and environment and more. Subjects in Transactions of the CSAM have been embodied by many internationally well-known index systems, such as: EI Compendex, CA, CSA, etc.

   下载PDF阅读器