Diagnosis method of water supply network leakage events based on time series shape analysis

Abstract

The invention provides a water supply pipe net leakage accident diagnosis method based on the time sequence shape analysis. The method comprises the following steps that firstly, a time sequence is constructed through water supply pipe net monitoring data, and the forward offset and the backward offset of the time sequence are determined; secondly, by utilization of the monitoring data, a historical subsequence base is constructed on the basis of the forward offset and the backward offset; thirdly, shape analysis is conducted on historical subsequences according to the cosine included angle distance between the historical subsequences, shape abnormality subsequences in the historical subsequence base are deleted, and a reference subsequence base is constructed; and fourthly, new subsequences are constructed, the reference subsequence base is compared with the new subsequences in shape, and leakage accidents are diagnosed according to the comparison result. According to the water supplypipe net leakage accident diagnosis method based on the time sequence shape analysis, accurate judgment of the leakage accidents in the water supply pipe network is achieved.

Description

technical field [0001] The disclosure belongs to the field of urban water supply network and data analysis, and specifically relates to a method for diagnosing leakage events of a water supply network, in particular to a method for diagnosing leakage events of a water supply network based on time series shape analysis. Background technique [0002] In 2016, the leakage rate of my country's urban water supply pipe network was about 18%, compared with the goal of the leakage rate of public water supply pipe network not exceeding 10% by 2020 as required by the "Water Pollution Prevention and Control Action Plan" ("Water Ten Measures") , there is still a large gap. The life cycle of the water supply network is very long, and the service period can reach 50 or even 100 years. According to statistics, my country's existing water supply pipe network is about 500,000 km, and these pipe networks will gradually enter the high-collar service period. It is estimated that by 2030, about...

AI Technical Summary

Problems solved by technology

However, the training of classification models requires labeled data (both normal and leaky), which belong to supervised learning methods

Leakage data in the water supply network is scarce, and due to management issues, water companies do not record all leakage events, which creates obstacles for the establishment of classification models

Images