基于神经网络深度学习和结构优化的供水管网漏口失效形式及漏失量预测

为了量化供水管网漏失危害,根据漏失形式探究管网漏失原因,在管网漏失试验的基础上,通过分析不同管材、管径、漏口失效形式的管段在各种压力下漏失瞬态漏口周围20个压力波动数据,利用BP神经网络在压力数据与漏失量、漏口失效形式之间建立映射关系,并通过大量训练来优化BP神经网络结构。结果表明,优化后的神经网络可精确地预测出管网漏失量及漏口失效形式,为一种适用性强、快速有效的预测方法。

Leakage Failure Form and Leakage Prediction of Water Supply Network Based on Neural Network Deep Learning and Structure Optimization

In order to quantify the harm of leakage in water supply network, based on the leakage experiment of the pipe network, through the analysis of 20 pressure fluctuation data around the leakage transient leakage of different pipe, pipe diameter and leakage failure forms under various pressures, the BP neural network is used to establish the mapping relationship between the pressure data, the leakage amount and the leakage failure form. The BP neural network structure is optimized through a lot of training. The experimental results show that the optimized neural network can accurately predict the leakage amount and the leakage failure form of the pipe network, which is a highly applicable, fast and effective prediction method.