Grade diagnosis of leakage faults in actual heating network based on improved CUSUM algorithm

In order to diagnose leakage faults in heating networks timely and effectively and improve the search efficiency of leakage locations in networks, based on the improved CUSUM algorithm, this paper presents a method for leakage grade diagnosis through the free combination of network topology structure. Taking the central heating network of an urban area as the research object, it is divided into five topology structures and their respective hydraulic condition models of the network are established. Four one-point leakage conditions and four two-point leakage conditions are selected for verification and analysis. The results show that under the condition that the absolute values of the maximum relative errors between the simulated values and the actual operation data of pressure and flow of the heating network leakage diagnosis model do not exceed 10%. The diagnosis accuracy rate of the leakage location of this method reaches 75%, verifying the feasibility of this method applied to the actual network.