基于神经网络的污水处理厂水质预测模型

Prediction Model Based on Neural Network for Water Quality of Sewage Treatment Plant

该文针对污水处理厂进水水质变化及处理过程中具有非确定特性、不易精确控制等特点,充分利用神经网络模糊信息处理的优势,结合污水处理工艺过程,以MATLAB为研究平台建立一套具有3层神经网络结构的预测模型并进行仿真验证。结果表明该系统能够预测该污水处理厂12 h后处理水质的状况,为运行管理提供可靠的数据支持,是一种预警处理水质的有效方法。

Abstract Considering uncertainty and difficulty to control accurately of influent quality during wastewater treatment, a 3-layer neutral network prediction model was established using MATLAB language on the basis of fuzzy processing of neutral network. The resuh shows this prediction model is able to predict 12-hour later water quality of treatment plant which can help superintendence master reliable data. It is an effective way to predict and forecast water quality for application.