供热;二次网;泄漏故障诊断;流量监测点;FAST-LTS算法;CUSUM方法

This study develops a leakage fault diagnosis algorithm for central heating systems to improve the leakage fault diagnosis efficiency and reduce the heating secondary network water loss rates. The method integrates the CUSUM (cumulative sum) change-point theory with the FAST-LTS algorithm from least trimmed squares (LTS) estimation as the decision criterion, combined with a pipe network simulation model. Taking the heating secondary network (single heat source branched network) of a residential area in Shandong Province as the research object, the proposed algorithm is preliminarily validated under three leakage rate conditions with the same topology structure and different leakage positions in the experimental pipeline network. Then, the actual flow data is collected using an ultrasonic flowmeter to validate the simulation model of the pipeline network. Finally, the parameters (e, o) are empirically determined to be (1.015, 0.6σ), and the flow monitoring points replace the traditional pressure monitoring points to overcome limited pressure data availability at heat sources. The research results show that for the diagnosis of four leakage conditions, when the error distance L is 20 m, the accuracy of leakage fault diagnosis reaches 75%, which verifies the feasibility and applicability of the algorithm proposed in this paper.