分析了BP神经网络原理及污水处理过程中BP网络模型设计,经过对苏州市某污水厂123组实测数据异点剔除和归一化处理后,对网络进行了训练和学习,30组数据用来检验网络,并用另外70组实测数据进行了实际仿真模拟,分析模拟结果表明,对各出水水质指标的模拟平均相对误差均小于5%,建立BP神经网络活性污泥法污水处理过程模型是有效的、可行的。%The paper analyzes the BP neutral network principle and BP network model design in the drainage treatment process,trains and learns the network after the different points rejecting and normalization processing of 123 group tested data of some drainage plant in Suzhou,adopts 30 groups to test the network and undertakes the simulation of the tested data of other 70 groups,and proves by the results of the simulation and results that the average relative error of the simulation of the water quality indexes is less than 5%,so it is effective and feasible to establish the model of the BP neutral network＇s activated sludge process to treat the sewage.
展开▼
