土壤成分与特性参数光谱快速检测方法及传感技术  

Fast Measurement and Advanced Sensors of Soil Parameters with NIR Spectroscopy

作　　者：李民赞[1] 郑立华[1] 安晓飞[1] 孙红[1] 

机构地区：[1]中国农业大学现代精细农业系统集成研究教育部重点实验室,北京100083

出　　处：《农业机械学报》2013年第3期73-87,共15页Transactions of the Chinese Society for Agricultural Machinery

基　　金：'十二五'国家科技支撑计划资助项目(2011BAD21B01);国家自然科学基金资助项目(61134011)

摘　　要：近红外光谱技术在分析土壤成分含量以及理化特性参数方面获得了良好的预测精度。人工神经网络、遗传算法、小波变换和支持向量机等现代数据处理算法的应用,最大限度消除了光谱外界干扰、提取了光谱有效信息,使得土壤特性参数预测分析模型更准确、稳定。在进行土壤参数原位实时光谱检测时,如何消除土壤含水率、土壤粒度等的影响,还需要技术突破。开发便携式或车载式农田土壤光谱实时分析仪,是促进精细农业实践的重要措施,已开发的车载式土壤在线光谱仪可以实现多个土壤参数的分析,并达到了相当高的精度。进一步开发多功能土壤在线检测系统,利用土壤介电特性或机械特性与光谱特性测量结果相互补偿与校正以消除误差并提高测量精度,是未来的发展方向之一。光声光谱、激光诱导击穿光谱和太赫兹光谱技术等现代光谱分析方法在土壤成分与特性参数分析方面表现出很强的能力,开展基础研究,揭示这些光谱技术在不同土壤类型、不同土壤成分条件下的吸收特征参数,是未来的研究方向。Near infrared spectroscopy （NIRS） is a powerful tool in the analysis of the soil compositions as well as soil physical and chemical properties. Artificial neural networks （ANN）, genetic algorithm （GA） , wavelet transform, support vector machine （SVM） and other modern data processing algorithms have made estimation and analysis of the soil parameters more accurate and stable since they can eliminate spectral error caused by outside interference and efficiently extract spectral information. It is still necessary to eliminate the impact of soil moisture and particle size on soil spectra in the in-site measurement. Development of the portable and vehicle-mounted real-time soil spectral analyzer is an important measure to promote the practice of precision agriculture. The developed vehicle-mounted soil on-line spectrophotometers have realized the analysis of several soil parameters, and reached a very high accuracy. It is a trend to develop multifunctional vehicle-mounted real-time soil spectral analyzer in the future by which the correction of the spectra can be performed with the data fusing of the spectral data and other dielectric or mechanical properties. The modern spectroscopies, photoacoustic spectroscopy （PAS）, laser induced breakdown spectroscopy （LIBA）, and terahertz （THz）, have shown the great potential in the analysis of soil compositions and parameters. It is needed to carry out basic researches to reveal the spectroscopy characteristics of soil samples in different types and conditions.

关 键 词：土壤 近红外光谱 土壤传感器 精细农业 

分 类 号：O657.3[理学—分析化学] S126[理学—化学]