| With the continuous expansion of the city gas pipeline construction. Pipeline in service of the accident frequency also always increase. In order to protect the safe operation of the pipeline, people’s lives and properly are threatened, to strengthen the management of the natural gas pipeline and to defend, indemnify along people’s lives and property safety, to ensure the security and stability of the pipe run have been inevitable.Based on the present situation of city gas pipeline, based on risk management are put forward on the basis of the principal component analysis and the BP neural network of comprehensive evaluation thoughts, principal component analysis is the variables of the complex relationship between interrelated simplified analysis method, in order to fully system research questions will be considered a lot of technical and economic index, principal component analysis (data information lost trying to at least under the principle of high dimensional variables of space dimension reduction processing, not only simplify the neural network training process, and improve the working efficiency and the accuracy of the results. And the BP neural network has parallel processing, adaptive learning, fault tolerance, good etc, and therefore can build a based on principal component and the BP neural network model is realized, and the risk of gas pipeline in effective evaluation.Based on this, it further use matlab software to realize the BP neural network analysis, sample training process that graph, from figure out can see network the output value and expected very close, and used the2groups of test data for testing, the results the error than traditional neural network method are much smaller, prove the feasibility of this method.At the end of the article mainly introduces the urban geographical information system (GIS) and its conventional function, can realize the basic of city gas pipeline for dynamic management and synchronous database building, and improve the management efficiency. For related personnel to provide scientific management decision. |