摘要
A new simulation strategy is proposed for the starting process of missile turbojet engine windmill. The starting process of windmill before ignition is simulated using a radial basis function neural network (RBFNN) , and the acceleration process after ignition which model is a set of nonlinear equations is solved using a particle swarm optimization (PSO) algorithm. The introduction of PSO helped to tackle the problem of divergence caused by traditional iteration methods. The calculated result is in a great agreement with test data, which shows that the presented model has a high accuracy. The starting processes are simulated at different ignition times, and the results are analyzed synthetically. The analysis shows how the ignition time affects the starting performance of engine windmill. The method offers a useful tool for ignition time optimization as well as engine starting performance analysis.
A new simulation strategy is proposed for the starting process of missile turbojet engine windmill. The starting process of windmill before ignition is simulated using a radial basis function neural network （RBFNN） , and the accelera- tion process after ignition which model is a set of nonlinear equations is solved using a particle swarm optimization （PSO） algorithm. The introduction of PSO helped to tackle the problem of divergence caused by traditional iteration methods. The calculated result is in a great agreement with test data, which shows that the presented model has a high accuracy. The starting processes are simulated at different ignition times, and the results are analyzed synthetically. The analysis shows how the ignition time affects the starting performance of engine windmill. The method offers a useful tool for ignition time optimization as well as engine starting performance analysis.
基金
Sponsored by the National Aeronautical Science Foundation of China(20095584006)
