基于遗传算法近红外光谱检测土壤养分的研究 |
| |
引用本文: | 刘雪梅,章海亮. 基于遗传算法近红外光谱检测土壤养分的研究[J]. 灌溉排水学报, 2013, 32(2) |
| |
作者姓名: | 刘雪梅 章海亮 |
| |
作者单位: | 华东交通大学土木建筑学院,南昌,330013 |
| |
基金项目: | 国家自然科学基金项目,2010年博士点基金批准项目,江西省科技支撑项目 |
| |
摘 要: | 利用可见/短波近红外光谱(Vis/SW-NIRS)分析测量土壤速效氮(N)和速效钾(K)。探讨遗传算法在分析测量土壤养分上的应用,根据优化结果采用最小二乘支持向量机(LS-SVM)方法建立校正模型。结果表明,LS-SVM模型优于PLS模型;GA-LS-SVM模型预测速效氮的精度较高。基于遗传算法可见/短波近红外光谱利用LS-SVM建模,可以作为一种土壤理化性质的测定方法。
|
关 键 词: | 遗传算法 近红外漫反射光谱 土壤 速效氮 速效钾 |
Study on Soil Nutrients by Using Near-Infrared Spectroscopy Based on the GA method |
| |
Abstract: | Visible infrared spectroscopy(Vis/SW-NIRS) was investigated to study the measurement accuracy of soil available nitrogen(N) and available potassium(K).Correction model was established using LS-SVM according to the optimal results in order ot explore the application of genetic algorithm in measuring soil nutrients.The results indicated that,LS-SVM models outperformed PLS models;Compared with available K,forecasting precision of GA-LS-SVM models was more preferable for available N.Visible and short wave-near infrared spectroscopycombined with LS-SVM based on GA could be utilized as a method for the determination of soil properties. |
| |
Keywords: | genetic algorithm(GA) Vis/SW-NIRS soil available nitrogen available potassium |
本文献已被 万方数据 等数据库收录! |
|