摘要
针对活性污泥污水处理过程强非线性、不确定性和时变性等综合复杂特性,本文研究了污水水质COD的检测技术现状,并总结了现有水质软测量模型的存在问题,最后建立适应大范围变化工况条件、并具有较高精度的多模型结构的水质软测量模型。其中,工况识别机制采用案例推理技术实现,每个局部模型采用机理模型结合RBF神经网络技术实现。
Concerning the complex characters of the wastewater treatment with activated sludge process, i.e., nonlinear, uncertainty and time-varying, etc. This paper discusses current water quality COD detection technology, and existing problems of water quality soft sensor. The water quality soft sensor model, based on multi-model con- struction, is of high precision and adapts to changing conditions. The condition recognition mechanism is realized by case-based reasoning technology, and the partial models are completed by mechanism model and RBF neural network technique.
出处
《科技广场》
2013年第9期71-75,共5页
Science Mosaic
关键词
污水处理
水质模型
软测量
神经网络
Wastewater Treatment
Water Quality Model
Soft Sensor
Neural Network
