The Research On Adaptive Control Technology For Micro Turbine Engine

The technology of Full Authority Digital Electronic Control (FADEC) makes it possible to adapt the advanced control strategy in aero-engine control systems. Single neuron adaptive control and adaptive control based on neural network prediction have been studied for the micro turbine engine by theory simulation, hardware-in-loop simulation and rig test.The control system hardware-in-loop simulation environment is established in tow aspects. Firstly, based on the Complex Programmable Logic Device and SOC microcontroller C8051F021, the signal emulator is developed to provide the interface signal for the simulation system. Secondly, a mathematical model of the micro turbine engine, which includes starting behavior model, is developed to provide the simulation environment for the study of adaptive control algorithm.According to the wide operating condition range and unstability of the fuel pump of the micro turbine engine, a kind of adaptive PID controller based on single neuron is applied, where the PID parameters can be self-adjusted by the error, error rates and the accelerator of error. This control algorithm has been realized in the electronic control unit. Hardware in-loop simulation and rig test results show that, the controller has good dynamic and static performance, meanwhile, satisfys realtime, and can be used in the electronic control unit.In order to solve the time delay problem, an adaptive control method based on the RBF neural network prediction was applied in micro turbine engine control system. The results of hardware-in-loop simulation and digital simulation indicate that: forecasting model can accomplish the prediction of rotational speed, and effectively overcome the time delay problem in control system.