APP下载

基于GA?BP神经网络的城市用水量预测

2020-08-03武风波赵盼吕茜彤

现代电子技术 2020年8期

武风波 赵盼 吕茜彤

摘  要: 城市用水准确的预测结果,对城市供水系统的控制具有直接的影响,而良好的城市用水控制系统不仅能够提高城市各个时段的供水效率,而且对城市人民生活幸福指数有较高的影响。该文设计通过分析传统的BP神经网络对城市用水量预测容易陷入局部误差极小,预测结果存在一定误差,提出在BP神经网络的基础上通过遗传算法优化BP神经网络进行城市用水量预测。通过设计GA?BP神经网络的具体结构,对已知的城市每日时用水量数据进行网络训练和学习。结果显示该模型具有一定的精度和适用性,预测结果可用于城市供水优化调度模型。

关键词: 城市用水; 用水量预测; BP神经网络; 预测建模; 网络训练; 仿真分析

中图分类号: TN711?34; TP872                   文献标识码: A                      文章编号: 1004?373X(2020)08?0147?04

Urban water consumption prediction based on GA?BP neural network

WU Fengbo, ZHAO Pan, L? Xitong

(School of Communication and Information Engineering, Xian University of Science and Technology, Xian 710054, China)

Abstract: The accurate prediction of urban water usage has a direct impact on the control of urban water?supply system. The good urban water control system can not only improve the water?supply efficiency in every time bucket of the city, but also have a serious impact on the urban people′s life happiness index. On the basis of the analysis that the traditional BP neural network is easy to fall into the local error minimum in the prediction of urban water consumption, and the prediction results has a certain error, the urban water consumption prediction based on BP neural network optimized by genetic algorithm is proposed. The specific structure of GA?BP neural network is designed to perform the network training and learning of known data of water consumption per hour in a city. The results show that the model has a certain accuracy and applicability, and the prediction results can be used in the urban water?supply optimization scheduling model.

Keywords: urban water usage; water consumption prediction; BP neural network; prediction modeling; network training; simulation analysis

0  引  言

随着城市人口的不断增加,城市供水系统也逐渐显示出它的不足,例如在用水高峰期部分高层供水水压较低,不能满足人们基本的生活需求。而我国传统的供水调度系统也存在较大的问题,例如水资源调度消耗能源较大,水资源浪费较多等。因此,为了满足城市人口的需求,必须对用水系统进行优化,而优化的前提是需要进行精确的用水量预测。

我国用水量的预测研究整体处于较低的水平,从最开始以每月用水量数据为研究依据到现在以每天日时用水量以及影响因素为研究依据,通过逐渐采用智能算法对其预测[1]。通过分析几种预测模型,神经网络算法相较于其他有更好的预测效果,故选用神经网络进行用水量预测。……

登录APP查看全文