基于混合模型的污水处理过程建模的研究  

Research on modeling of wastewater treatment process based on hybrid model

作　　者：史鹏涛 王璟德[1] 王健红[1] SHI Peng-tao;WANG Jing-de;WANG Jian-hong(College of Chemical Engineering,Beijing University of Chemical Technology,Beijing 100029,China)

机构地区：[1]北京化工大学化学工程学院,北京100029

出　　处：《计算机仿真》2020年第8期188-191,408,共5页Computer Simulation

摘　　要：针对单纯应用ASM系列机理模型或神经网络模型在污水处理过程建模中适用性差、难以应用的缺点,采用ASM2简化的SPM机理模型与BP神经网络相结合的混合模型建立了活性污泥污水处理过程模型,通过BP神经网络不断校正机理模型参数,使得参数随污水入口组成的变化自适应调整,最终得到了性能可观的混合模型,并分析了混合模型对工业污水处理过程的适应能力。仿真结果表明,混合模型能够克服传统模型的不足,能够通过调节模型参数改善预测结果,优势在于能更好的利用数据中蕴含的因果关系。Aiming at the shortcomings of using ASM series mechanism model or neural network model alone in the process of wastewater treatment modeling,which is poorly applicable and difficult to apply,a model of activated sludge wastewater treatment process was established by using a hybrid model of combining ASM2 simplified SPM mechanism model and BP neural network.The parameters of the mechanism model were continually corrected using the BP neural network,so that the parameters were adaptively adjusted with the change of the sewage inlet composition.Finally,a mixed model with considerable performance was obtained,and the adaptability of the mixed model was analyzed for the industrial wastewater treatment process.The simulation results show that the hybrid model can overcome the shortcomings of the traditional model and can improve the prediction results by adjusting the model parameters,and the advantage is that it can make better use of the causal relationship contained in the data.

关 键 词：机理模型 神经网络 污水处理 建模 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]