神经网络在工业污水处理建模中的应用

研究工业污水指标在线准确检测问题,工业污水处理是一种复杂的动态生物反应工程系统,具有多变量、非线性等特点,难以建立精确的数学模型,且在线实时检测仪表价格昂贵。从应用工程角度出发,提出将遗传神经网络方法应用于污水指标实时检测,建立污水指标软测量模型。利用一些可测指标与水质指标间的关联关系,通过神经网络进行拟合,并采用遗传算法优化神经网络参数。仿真结果表明,建立的神经网络污水估计模型,可以对污水指标进行实时检测和估计,并为污水在线实时检测提供了新的思路。

Research on Industrial Sewage Disposal Model Based on Neural Network

Industrial sewage disposal is a complex and dynamic biological reaction engineering system.It is a multi variables and nonlinear system,for which it is difficult to establish a precise mathematical model,and the on-line real-time detecting instrument is expensive.From the application point of view of engineering,the genetic neural network method was applied to the real-time detection of sewage indicators and establish sewage indicators soft measurement model.Using some measurable indicators and the indicators of water quality between the associations,and through the neural network fitting,the genetic algorithm was used to optimize the parameters of neural network.The simulation results show that,the estimate mode for sewage disposal based on neural network model can detect and estimate the sewage indicators in real-time,and it provides a new thought for the on-line and real-time detection for sewage water.