基于模糊算法的多小脑神经网络在混凝投药系统中的应用

APPLICATION OF FUZZY ALGORITHM-BASED MULTIPLE CMAC NEURAL NETWORKS IN COAGULANT DOSING SYSTEM

净水工艺中的混凝投药是一个受多变量影响的非线性系统,无法建立准确的数学模型,且滞后时间较长。为了将混凝投药后的出水浊度控制在设定的范围内,并且有效地减少投药量,通过对混凝投药过程的分析,设计一种基于模糊算法的多小脑神经网络(CMAC神经网络)前馈控制器,并设计控制器离线建模和在线学习的方法。最后使用MATLAB进行仿真验证,结果表明,该前馈控制器能够在原水浊度和原水温度变化的情况下,有效地将浊度控制在设定的范围内,并且能够实现投药量的优化。

Coagulant dosing of water purification process is a nonlinear system affected by multiple variables, it is difficult to build accuratemathematical model, and has quite long time delay. In order to control the coagulant dosed effluent water turbidity within the set range and effectivelyreduce dosage,we designed a fuzzy algorithm-based multiple C M A C neural networks feed-forward controller,which is based on the analysisof the process of coagulant dosing process. Moreover,we also designed the methods of off-line modelling and online learning. Finally,weused M A T L A B for simulation verification,result showed that the feed-forward controller could effectively control the turbidity within the setrange under the condition of raw water,s turbidity and temperature all changing,and the dosage could be optimised as well.