Research On The Detection Method Of Oil Pipe Network Leakage Based On Fuzzy Neural Network

Pipeline transportation which is irreplaceable has been playing an increasingly important role in the development of the national economy. However, with pipelines increasing and time passing by, pipe network leakage accident due to corrosion, natural disasters and other factors occurs frequently. Therefore, the research on the detection method of oil pipe network leakage is of great significance.It is difficult to satisfy the requirements of precision and accuracy by using the traditional detection method. And because the pressure curve caused by device adjustment in monitoring station is similar to that caused by pipe network leakage, traditional detection method brings about some errors. For uncertainty of pipe network’s running state and complexity of device adjustment, detection method based on fuzzy neural network is proposed in this paper. The following three issues are solved:data acquisition method, query method of device adjustment in monitoring station and detection method of pipe network leakage.Firstly, OPC data acquisition client of pipe network is designed. This system, which has real-time reliable performance, is easy to maintain and extend. This method has many advantages compared with conventional driver mode and dynamic data exchange mode. In order to collect pressure, flow, density signal and information of device adjustment, OPC client software is designed from two aspects of functionality and modules. And specific steps to develop the software in Visual C++ environment are discussed in detail. And the objectives of data acquisition and supervisory control are achieved.Secondly, in order to obtain normal adjustment information causing pressure of detecting station change, search method for traversing monitoring stations in pipe network is proposed. From the points of time complexity and space complexity, two traversal methods are researched respectively:depth-first search and breadth-first search. Ideas and processes of the two algorithms are discussed in detail. On this basis, in order to improve real-time performance of detection system, an improved algorithm of breadth-first search is put forward to satisfy demand of levelorder traverse to query device adjustment in monitoring station.Finally, detection method of oil pipe network based on fuzzy neural network is presented in this pape. Fuzzy neural network combines the advantages of fuzzy system and neural network. Fuzzy rules can be obtained by learning. Detection method based on type-1 fuzzy neural network and interval type-2 fuzzy neural network are studied in this paper by introducing structures and learning methods. Following conclusions are reached by simulation: Detection method based on interval type-2 fuzzy neural network has higher accuracy, but it needs more time for training.