基于小波神经网络的污水处理污泥体积指数软测量

Sludge volume index soft sensing based on wavelet neural network for sewage treatment

污水处理过程具有非线性、时变和滞后等特点,因而无法进行准确的数学建模。现有的污水处理技术中最突出的问题是一些关键的水质参数不能在线监测,只能通过人工间接测量再通过计算获得,耗时较长,不能及时地进行信息反馈,会造成一些严重的后果。为了避免这样的问题,提出了基于小波分析的神经网络(BP)软测量技术,通过建立小波神经网络参数软测量模型,对污水处理中难测水质参数SVI(污泥体积指数)进行在线监测。研究表明,此方法能有效规避单一的BP算法收敛速度慢、容易陷入局部最优解等问题,有助于实现对污水处理的智能控制。

Since the process of wastewater treatment has the characteristics of nonlinearity, time-varying and hys- teresis, it is impossible to carry out accurate mathematical modeling. The most prominent problem is that some key water quality parameters cannot be monitored on-line in present wastewater treatment technology. It can only be monitored through manual and indirect measurement and calculation, which takes a long time and cannot pro- vide timely information feedback, causing a number of serious consequences. In order to avoid such problems, a neural network （BP） soft sensing technique based on wavelet analysis has been proposed. The difficult monitored water quality parameters of SVI （sludge volume index） in wastewater treatment are monitored on-line through the establishment of parameter soft-sensing model of wavelet neural network. The research shows that this method can effectively avoid the problems in single BP algorithm, such as slow convergence speed, easy to fall into local opti- mal solution, and so on, which is helpful to realize the intelligent control of wastewater treatment.