论文部分内容阅读
针对污水处理过程具有非线性的特点,建立基于PSO-ESN神经网络的污水处理软测量模型,来对于污水处理关键水质参数BOD(Biochemical Oxygen Demand)进行预测。由于回声状态网络(Echo State Network,ESN)学习算法无法有效解决高维矩阵训练不可逆,采用基于粒子群优化算法对于回声状态神经网络输出权重进行训练,进而有效解决回声状态网络病态解的问题。仿真结果证明,所设计的基于关键水质参数生化需氧量(BOD)软测量模型,其应用在污水处理关键水质参数预测的有效性,且该软测量模型具有较高测量精度。
In view of the non-linear characteristics of sewage treatment, a soft water measurement model of sewage treatment based on PSO-ESN neural network is established to predict the critical water quality parameter BOD (Biochemical Oxygen Demand). Because Echo State Network (ESN) learning algorithm can not effectively solve the irreversibility of high-dimensional matrix training, particle swarm optimization algorithm is used to train the output weights of echo state neural networks, so as to effectively solve the problem of echo state network ill-posed. The simulation results show that the designed BOD soft-sensing model based on the key water quality parameters is effective in predicting the critical water quality parameters of wastewater treatment, and the soft-sensing model has high measurement accuracy.