Safe Intelligent Control Method with Limited Tracking Errors of Aeroengines

In the case of command tracking, traditional control algorithms such as PID, due to the limitations of the algorithm itself, for complex control systems such as aeroengines, it is impossible to make clear safety limits or processing for the errors of command tracking in many cases. For the above design defects, this paper adopts safety reinforcement learning with obstacle function combined with errors state space model to limit the tracking errors and explore the optimal control strategy at the same time. For the aeroengines control system with limited tracking errors, the proposed method that is a safety intelligent control method based on the barrier function, can limit the low-pressure rotor speed of the aeroengines within a given error range. Finally, the experimental results with the traditional PID controller show that the method has good trajectory tracking effect and strong anti-interference ability.