PSO-RBF神经网络在供水管网节能上的应用  

Application of PSO-RBF neural network in energy saving of water supply network

作　　者：王力[1] 孙贺 WANG Li;SUN He(College of General Aviation,Civil Aviation University of China,Tianjin 300300,China;College of Electronic Information and Automation,Civil Aviation University of China,Tianjin 300300,China)

机构地区：[1]中国民航大学通航学院,天津300300 [2]中国民航大学电子信息与自动化学院,天津300300

出　　处：《现代电子技术》2020年第1期136-139,共4页Modern Electronics Technique

基　　金：国家自然科学基金委员会与中国民用航空局联合资助项目（U1733119）;国家自然科学基金项目（U1333111）;中央高校基本科研业务费项目（3122017018,3122015D008）;中国民航大学专项资助（3122017015）

摘　　要：目前供水管网系统中的水泵机组存在用水规模大、过压供水的情况,为解决供水管网系统电量消耗大的问题,提出一种基于正则化RBF神经网络和粒子群算法的水泵压力控制策略。首先,通过供水管网结构分析寻找各区域最不利出水点;然后,利用历史数据训练正则化RBF神经网络使其具备水力模型辨识能力,并通过粒子群算法对神经网络学习过程进行优化,提高学习效率;最后根据当前压力输入输出样本集进行水泵压力自校正控制,实现水泵压力的智能控制。经过验证,该系统可以实现在满足供水管网用水需求的前提下,降低水泵压力,消除过压供水,有效地节约能源。At present,large water consumption and overpressure water supply occur to the pump unit of the water supply network system.In order to deal with the large power consumption in the water supply network system,a control strategy for water pump pressure based on regularized RBF neural network and particle swarm optimization is proposed.Firstly,the most unfavorable water outlet in each region is found by analyzing the structure of water supply network.Then,the regularized RBF(Radial Basis Function)neural network is trained by the historical data to get the ability of hydraulic model identification,and the learning process of neural network is optimized by the particle swarm optimization to improve the learning efficiency.Finally,the self⁃correction control of pump pressure is realized according to the current pressure input and output sample set to achieve the intelligent control of pump pressure.It is verified that the system can reduce the pump pressure,eliminate overpressure water supply and realize energy saving effectively on the premise of meeting water demand of water supply network.

关 键 词：供水管网节能 智能控制 RBF神经网络 粒子群算法 自校正控制 节能减排 

分 类 号：TN911.1-34[电子电信—通信与信息系统] TP183[电子电信—信息与通信工程]