TNT生化降解时间序列的人工神经网络预报模型

Artificial Neutral Network Predicting Model of TNT Biodegradation Time Series

利用变步长ＢＰ 算法，对白腐真菌生化降解实际ＴＮＴ 装药废水过程中ＴＮＴ 含量变化的时间序列建立了人工神经网络预报模型，并利用该模型对生化降解过程的变化规律及趋势进行了研究。结果表明，模型的计算值与实测值之间的误差很小，对未来时刻数据的预测精度也较高，模型较好地反映了ＴＮＴ 含量的变化规律。

An artificial neutral network model was established based on the time series of TNT concentration data coming from TNT packing wastewater biodegradation process by white rot fungi, using the varied pace back propagation (BP) algorithm. The model was then used in the study on the changing rule of TNT concentration and the developing trend. The results showed high accuracy both for the present data and for the predicting data, which showed that the established ANN model had reflected the rule of TNT packing wastewater biodegradation process. The study proved that ANN is a novel and proper approach for the study of TNT biodegradation.