BP神经网络对管网安全使用期的预测  

Research on Prediction of Safe Using Dates of Water Supplying Pipes Based on BP Neural Network

作　　者：阎立华[1] 马婷婷 田昊平[1] 

机构地区：[1]沈阳建筑大学市政与环境工程学院,辽宁沈阳110168

出　　处：《沈阳建筑大学学报（自然科学版）》2008年第5期845-849,共5页Journal of Shenyang Jianzhu University：Natural Science

基　　金：建设部科技计划(06-K4-14)

摘　　要：目的对管网安全使用期进行科学预测,为更换、检修、维护供水管网提供参考依据.方法采用BP人工神经网络的方法对管网安全使用期进行预测,以影响管网使用的因素作为输入变量,以管网使用时间作为输出变量,建立BP神经网络模型,从而实现对管网使用期的预测.结果经过17次训练,模型达到了要求的精度,相对误差均小于±10%.结论用BP神经网络可以对管网的安全使用期进行预测.In order to predict the safe using dates of water supplying pipes, and then offer a reference for changing, examining and repairing the water supplying pipes, consider establishing a BP neural network to predict. To make these factors which work on the safety of water pipes as the input variables, the using dates as the output variable, and then establish the BP network module, and predict the safe using dates. After being trained for 17 times, the module gets the required precision and the error for predict is under ± 10 %. In conclusion, BP neural network can be used in prediction of safe using dates, and the accuracy and practicability of the module are satisfing.

关 键 词：供水管网 安全使用期 BP神经网络 预测 

分 类 号：TU991.6[建筑科学—市政工程]