Research On The Diagnosis Of Urban Gas Pipe Network’s Leakage

The vigorous development of city gas has played an important role in the improvement of national economy and social undertakings. However, with the laying of the gas network area, the attendant security risk has become increasingly acute. In recent years, due to the corrosion of pipe, aging, dug, natural disasters and other factors, the pipe network’s leakage frequently occurrs, the safety of the city gas network has been a great challenge. Therefore, in order to ensure the safe operation of city gas network, the diagnosis methods of urban gas pipe network’s leakage have been studied in this article.(1)Studied the leakage monitoring of city gas network, and proposed the SCADA system combinde with the onlie simulation system can monitor the leakage. The leakage of city gas network can be detected timely by comparing the collection data of the SCADA system and the simulation data of the online simulation system. Set the alarm threshold of leakage to solve the problem of false positibes and false negatives.(2)The number of SCDA system monitoring points of city gas network is too small, can not meet the leakage monitoring and diagnostic accuracy. So used the sensitivity method to add the monitoring points, and improved the accuracy of the leakage monitoring and diagnosis.(3) The BP neural network was used to identify the fault pipe of the city gas network, and broke the traditional methods by strict logical reasoning and accurate calculation of the conclusion of the problem. This method improved the efficiency of leakage detection.By taking a city gas pipe network’ for example, the leak identification results validate the feasibility of the BP neural network.(4) Fuzzy clustering method was used to partition the city gas network, and the number of pipe sections in each region was significantly less than the number of pipe sections of the entire pipe network. The partition of network improved the recognition accuracy and efficency of the BP neural network. The result of pipe network partitions show that the pipes may be not connected to each other in the same area, so the pipe network partition is a virtual partition of some kind of inner link, rather than space on the partition.(5)The model method was used to locate the leakage of city gas network, and solved the shortcomings of the BP neural network can only identify the leak pipe, but can not identify the leak point. Construct the leak model and used the model method to calculate the leak location. By taking a city gas pipe network’for example, the usage of model method in calculating the leak location was verified by the results.(6)Based on the above research, the diagnosis system of city gas pipe network’s leakage was designed by using the VC++and MATLAB language. The system provides a friendly interface, easy to operate, to avoid the burden of manual calculation, and improve the efficiency of leakage detection.