摘要
针对污水处理过程所具有的多变量、非线性和大时滞的特点,利用出水水质参数与多个可测过程参数间的相关关系,给出了基于RBF神经网络的出水水质参数预测模型。采用免疫粒子群优化算法来训练网络的隐层节点、径向基函数的中心点和网络权值。以反应时间、DO浓度、ORP和pH值作为输入参数,实现对COD,NH3-N,TP等水质参数的预测。仿真试验表明,该预测模型对污水处理出水水质参数COD,NH3-N,TP具有理想的预测效果。
A effluent quality parameters predictive model based on RBF neural network is proposed for the sequential wastewater treatment processes,regarding the characteristics of multivariable,nonlinear,big-lagged in the treatment process,and utilizing the relationship between effluent quality parameters and the measurable process parameters.The nodes of hidden layer,the centers of Radial Basis Function and network weights are trained by applying the immune particle swarm optimization algorithm.Reaction time,DO（dissolved oxygen）,ORP（oxidation reduction potential） and pH are selected as input parameters,which can realize the prediction of the effluent quality parameters of COD（chemical oxygen demand）,NH3 and TP（total phosphorus）.Simulation experiments reveal that the predictive model proposed has good predictive effect for the water quality indices COD,NH3 and TP in the wastewater treatment process.
出处
《微处理机》
2010年第2期75-77,81,共4页
Microprocessors
基金
国家自然科学基金(60774032)
教育部高等学校博士学科点专项科研基金(新教师基金课题)(20070561006)
关键词
污水处理
神经网络
免疫粒子群算法
预测
Wastewater treatment
Neural network
Immune particle swarm algorithm
Prediction
