孙亚楠,李仙岳,史海滨,崔佳琪,王维刚.河套灌区土壤水溶性盐基离子高光谱综合反演模型[J].农业机械学报,2019,50(5):344-355.
SUN Ya’nan,LI Xianyue,SHI Haibin,CUI Jiaqi,WANG Weigang.Hyperspectral Integrated Inverse Model for Water-soluble Salt Ions Content in Hetao Irrigation District[J].Transactions of the Chinese Society for Agricultural Machinery,2019,50(5):344-355.
摘要点击次数: 1539
全文下载次数: 756
河套灌区土壤水溶性盐基离子高光谱综合反演模型   [下载全文]
Hyperspectral Integrated Inverse Model for Water-soluble Salt Ions Content in Hetao Irrigation District   [Download Pdf][in English]
投稿时间:2018-10-31  
DOI:10.6041/j.issn.1000-1298.2019.05.039
中文关键词:  水溶性盐基离子  光谱变换  特征波段  特征光谱指数  支持向量机  综合反演模型
基金项目:国家自然科学基金项目(51539005、51669020、51469022)、内蒙古自治区水利科技重大专项(NSK2017-M1)和内蒙古自然科学基金项目(2016JQ06)
作者单位
孙亚楠 内蒙古农业大学 
李仙岳 内蒙古农业大学 
史海滨 内蒙古农业大学 
崔佳琪 内蒙古农业大学 
王维刚 内蒙古农业大学 
中文摘要:为了提高野外高光谱反演土壤水溶性盐基离子的精度,以河套灌区永济灌域盐渍化土壤为研究对象,构建了基于光谱变换、特征波段、特征光谱指数筛选以及支持向量机(SVM)的机器学习相结合的高光谱综合反演模型。结果表明,经预处理的原始光谱反射率与土壤离子相关性总体较低,最大相关系数仅为0.18,原始光谱反射率与土壤离子的相关系数由大到小依次为Ca2+、 SO2-4、 Mg2+、全盐量、Na++K+、Cl-。全盐量、Na++K+、Cl-、SO2-4、Ca2+、Mg2+的光谱最优变换形式分别为(1/R)″、(1/R)″、(lnR)′、(lnR)″、R′、(lnR)″,敏感波段(P<001)数分别为41、7、9、65、76、28个,利用逐步回归法在敏感波段中筛选出特征波段,基于特征波段建立的回归模型中各离子的决定系数R2平均值为035,均方根误差RMSE平均值为087g/kg,其中SO2-4拟合精度最高,R2为0.52,Ca2+拟合精度最低,R2仅为0.20。将特征波段代入光谱指数中,结合逐步回归法确定了Mg2+特征光谱指数为3个,全盐量特征光谱指数为2个,Na++K+、SO2-4、〖JP3〗Ca2+特征光谱指数分别为1个,与仅考虑特征波段的回归模型相比,特征波段+〖JP〗特征光谱指数结合后各离子回归模型的R2平均提高了58.67%,RMSE降低了24.60%,其中SO2-4拟合精度最高,R2为0.74,RMSE为0.47g/kg。考虑了特征波段+特征光谱指数的SVM模型相比仅考虑特征波段的SVM模型,其预测能力有了明显提高,各离子相对分析误差(RPD)平均提高了11027%,训练集R2平均提高了3754%,RMSE平均降低了4012%,验证集R2平均提高了56.04%,RMSE平均降低了39.39%。SO2-4的RPD达到3000,模拟效果最优,具备很好的预测能力;全盐量模型具有很好的定量预测能力,Mg2+模型可用于评估或相关性方面的预测,Na++K+、Ca2+的模型具有区别高低值的能力。
SUN Ya’nan  LI Xianyue  SHI Haibin  CUI Jiaqi  WANG Weigang
Inner Mongolia Agricultural University,Inner Mongolia Agricultural University,Inner Mongolia Agricultural University,Inner Mongolia Agricultural University and Inner Mongolia Agricultural University
Key Words:water soluble salt ions  transformation of hyperspectral  characteristic band  characteristic spectral index  support vector machine  integrated inverse model
Abstract:It is significant to take best agricultural measures and improve salinization to rapidly and accurately determinate the composition and content of soil salt. The hyperspectral integrated inversion model based on transformation of hyperspectral, characteristic bands, characteristic spectral indices screening and support vector machine (SVM) was established to improve the accuracy of water soluble salt ions content by taking the saline soil of Yongji irrigation area of Hetao Irrigation District. The results showed that the correlation between the original spectral reflectance by pretreatment and water soluble salt ions content was relatively low and the maximum correlation coefficient was 0.18, and the sequence of them from big to small was as follows: Ca2+, SO2-4, Mg2+, the content of salt, Na++K+ and Cl-. The optional transformation forms of salt content, Na++K+, Cl-, SO2-4, Ca2+ and Mg2+ were (1/R)″, (1/R)″, (lnR)′, (lnR)″, R′ and (lnR)″, respectively. The numbers of sensitive bands (P<0.01) were 41, 7, 9, 65, 76 and 28, respectively. Stepwise regression method was used to filtrate the characteristic bands from sensitive bands, and the average of determination coefficient (R2) and the average of root mean square error (RMSE) of each ion in the regression model based on the characteristic band were 0.35 and 0.87g/kg, of which R2 was the largest and the smallest were SO2-4 (0.52) and Ca2+ (0.20), respectively. Combined with the stepwise regression method, the characteristic bands were substituted into the spectral index to determine that there were three characteristic spectral indices for Mg2+, there were two characteristic spectral indices for salt content, and there were one characteristic spectral index for Na++K+, SO2-4 and Ca2+, respectively. The R2 of model for water soluble salt ions content based on the characteristic bands and characteristic spectral indices was increased by 5867%, and the RMSE was decreased by 2460%, of which the maximum R2 was SO2-4 (0.74), RMSE was 0.47g/kg. 〖JP〗The model of SVM based on the characteristic bands and characteristic spectral indices combined had a significant improvement in the prediction than that merely based on the characteristic bands, for example, the average relative analysis error (RPD) was increased by 110.27%, the R2 was increased by 37.54% and the RMSE was decreased by 4012% in the training set, the R2 was increased by 5604% and the RMSE was decreased by 3939% in the verification set. The results showed that the RPD of SO2-4 reached 3000, which showed a good prediction ability. The model of salt content and Mg2+ had good quantitative prediction ability which can be used for assessment or correlation prediction, respectively. The SVM models of Na++K+ and Ca2+ had the ability to distinguish between high and low values. 

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