Research on reliability simulation prediction of natural gas pipeline system

In order to master the running status messages of natural gas pipeline system, predict the possible time and which equipment of the system may be go wrong, and change the equipment in time, avoid the failure and economic losses caused by the system failure. so, the prediction of system reliability is very important to system manage and maintenance. For traditional reliability prediction method, such as BP, RBF and SVM, requires a large sample size and couldn't effectively forecast the factory chaotic nonlinear object is contained in a defect, is presented based on the geometry of chaos phase space reconstruction and information SVM to predict (CSVM) algorithm. It is beyond the reach of the traditional forecasting methods because of the chaotic nonlinear and large sample size. Chaos SVM Prediction was studied based on the geometric information, SVM and chaos theory. Natural gas pipeline Henon system was used to test the model in this report, application results show that: (1.) under the background of chaos, the algorithm is superior to traditional BP, RBF and SVM forecasting simulation methods such as accuracy, efficiency and the number of support vector forecasting performance; (2.) the predicted time of the algorithm has a big increase than BP, RBF and SVM, by about 16.5 %..