CAJ | 학술논문

　　黄瓜在温室栽培面积中占有很大的比重，监测温室黄瓜营养状态非常重要。因此，利用光谱分析技术对叶片所含的化学成分进行测量，主要包括建模与预测两步，采用基于BP算法的人工神经网络和支持向量机进行光谱的定量分析，探索基于光谱分析的整株光谱判断温室黄瓜的营养状态。研究结果表明，基于 BP 算法的人工神经网络和支持向量机所建立的回归模型，在一定程度上提高了模型的自预测能力和实际预测能力。
　　황과재온실재배면적중점유흔대적비중，감측온실황과영양상태비상중요。인차，이용광보분석기술대협편소함적화학성분진행측량，주요포괄건모여예측량보，채용기우BP산법적인공신경망락화지지향량궤진행광보적정량분석，탐색기우광보분석적정주광보판단온실황과적영양상태。연구결과표명，기우 BP 산법적인공신경망락화지지향량궤소건립적회귀모형，재일정정도상제고료모형적자예측능력화실제예측능력。
The monitoring of greenhouse cucumber nutrition status is very important account for a large proportion in the greenhouse area .It is contained modeling and prediction two-step by using spectral analysis to measure the chemical composition of leaves quantitative analysis of spectra is exercised by Artificial Neural Networks based on BP and support vector machine .The methods is probed in the Study of monitoring the greenhouse cucumber nutritional status based on canopy reflectance .The results showed regression model established based on BP algorithm artificial neural networks and support vector machine improved the self-predictive ability of the model and actual predictive ability to some extent .