| The modeling and real-time simulation controlling system for the aero-engine could possess the advantages of reducing engine testing expenses, shortening development period, decreasing the cost and avoiding the rig-test risk, whereas, Matlab has the advantages on graphical modeling and visualized simulation and supplies a platform for establishing a flexible, extensible and high visual aeroengine model and carrying out real time simulation.Key topics are merged into the mainstreams of the paper, namely the development of aero-engine above-idle state model and startup model, control algorithm and real time simulation procedure design based on xPC Target. Firstly, the above-idle state component model is established by means of Matlab/Simulink system, the steady model and dynamic model are accomplished by solving coupling equations governed by Newton-Raphson algorithm. Digital simulation is carried out in Simulink, and the real-time simulation is implemented in target-host mode based on xPC Target .Secondly, startup model for turbo-shaft engine is established. The low-speed characteristics are predicted by the stage-stacking method due to unavailable characteristics at low speeds, and the component-level startup model is developed with the consideration of the influence of combustion efficiency variation, heat transfer and pressure ratio distribution in gas generators. Startup performances are analyzed. Finally, PID control algorithm based on neural network is designed in the above-idle state model, the nonlinear mapping ability of neural network can realize the optimum choice of PID control parameters. The simulation is validated on xPC Target simulation platform.It shows that turbo-shaft engine model established in Matlab/Simulink and the research of real time simulation in xPC Target platform can satisfy the requirements of the real-time specification as well as the simulation precision, which open up a new approach for turbo-shaft control system design. |
PDF Full Text Request