摘要
测定了 47种取代苯类化合物在松花江水中的 5日生化需氧量 (BOD5) .分别采用线性基团贡献法和非线性基团贡献法(人工神经网络法 )对化合物的生物降解性BOD5 ThOD(ThOD :理论需氧量 )进行QSBR研究 .得到不同基团对生物降解性的贡献为 :C6H5>COOH >OH >CH3 O CH3 >NH2 >Cl >NO2 .线性基团贡献法对于训练组和测试组的定性预测正确率分别为72 %和 86 % ;而人工神经网络法的预测正确率分别为 92 %和 86 % .预测结果表明线性和非线性基团贡献法的预测效果都很好 ,相比而言 。
The biochemical oxygen demand (BOD 5) of 47 substituted benzenes were determined. The quantitative structure biodegradability relationship studies were performed with biodegradability (BOD 5/ThOD) by the linear group contribution method and artificial neural network approach, respectively. The contribution order of various group to biodegradation is C 6H 5>COOH>OH>CH 3O/CH 3>NH 2>Cl>NO 2. The correct prediction rate of the linear group contribution method is 72% for the training set and 86% for the testing set, while it is 92% and 86%, respectively, when by neural network approach. It has been shown that both linear and nonlinear group contribution method are able to fit very well for either set. However, nonlinear group contribution method can provide a superior fit to biodegradation and produce a lower prediction error than the linear group contribution method.
出处
《环境科学学报》
CAS
CSCD
北大核心
2002年第1期117-119,共3页
Acta Scientiae Circumstantiae
基金
国家自然科学基金 (2 98770 0 4)资助项目
