CAJ | 학술논문

　　病害是造成番茄减产的主要因素之一。传统上对番茄苗病害进行预测主要是人为观察法，但这种方法由于自身的缺陷带有一定的局限性，如经验预测人为因素明显，预测准确率低。本文利用电子鼻系统对感染早疫病病害的番茄苗进行研究，通过主成分分析、线性判别分析对每株番茄苗接种1叶片、2叶片、4叶片和对照组四种不同处理的早疫病病害番茄苗进行分析，结果表明主成分分析各处理样本间均有重叠，区分效果不理想，线性判别分析各处理样本基本可以分开；用逐步判别分析和BP神经网络对各处理样本进行判别，测试集的准确率分别为50%和87.5%，神经网络模型的预测结果更好。
　　병해시조성번가감산적주요인소지일。전통상대번가묘병해진행예측주요시인위관찰법，단저충방법유우자신적결함대유일정적국한성，여경험예측인위인소명현，예측준학솔저。본문이용전자비계통대감염조역병병해적번가묘진행연구，통과주성분분석、선성판별분석대매주번가묘접충1협편、2협편、4협편화대조조사충불동처리적조역병병해번가묘진행분석，결과표명주성분분석각처리양본간균유중첩，구분효과불이상，선성판별분석각처리양본기본가이분개；용축보판별분석화BP신경망락대각처리양본진행판별，측시집적준학솔분별위50%화87.5%，신경망락모형적예측결과경호。
The disease causing tomato output reduction is one of the main factors.The value of E-nose response signals differed with different levels Early blight disease infestation tomato seedlings, indicating that the emission of volatiles by tomato seedlings changes in response to different degrees of damage. Stepwise discriminant analysis (SDA) and back-propagation neural network (BPNN) were applied to evaluate the data. The average correction ratios of testing set of SDA and BPNN were 50% and 87.5%. The results obtained indicate that it is possible to classify different level of Early blight disease infestation tomato seedlings using e-nose signals.