Study On High Pressure Common Rail Fuel Injection System Performance Simulation And Evaluation Of Diesel Based On LS-SVM

The high-pressure common rail injection system has the characteristics：complex, nonlinearand time-varying. How to control and optimize the injection process, and improve theinjection performance of the high-pressure common-rail diesel effectively, were poor in theprocess of the study and development of the system. It is a primary task how to solve theproblems mentioned above for the experts and engineers. Therefore this paper studied on thesimulation of the high pressure common rail injection performance and its intelligenceevaluation based on the "985engineering" project II, Innovation platform of advance designand manufacture, science and technology for vehicle in Hunan university.(project of powertransmission and control). This topic is of theoretical significance and application value.Therefore, some methods such as AMESim software, least square support vector machinetechnology (LS-SVM), grey relation analysis and simulation experiment have been fused,which have been used to study the injection performance in high-pressure common rail modeland its intelligence evaluation. The main innovations and research work are expressed asfollows:(1) The working principle and the characteristics were studied. Based on common railsystem model built by using AMESim software, studied the effects of common rail pressure,common rail capacity, fuel injection pulse width, control chamber volume, needle springstiffness and preload force, pump speed, jet nozzle diameter and nozzle numbers on injectioncharacteristic，and also the impact of those parameters on rail pressure fluctuations.(2) Analyzed the influence of parameters on common rail injection performance furtherbased on grey relational theoretical, and calculated on the degree of influence to identify thekey structural which determine the input and output variables of the evaluation model. Thisstudy benefits the optimization and exploitation on common rail system.(3) Adaptive particle swarm optimization algorithm was used to optimize the LS-SVMparameters and evaluate the injection performance with the LS-SVM. Results show that, theevaluation accuracy is88.9%. The training time obviously reduced and it’s good forimproving and completing the common rail system.