An Aeroengine Control Method Based on Safe Reinforcement Learning

Aeroengine is a complex control system with strong nonlinearity. An aeroengine safe reinforcement learning control based on barrier function are proposed to improve the safety of aeroengine. Aiming at the safety limit and optimal control of aeroengine exhaust temperature (EGT), this paper takes the JT9D aeroengine on the T-MATS simulation platform as the experimental object. We can obtain the optimal control under the condition of satisfying the temperature safety limit by setting the barrier function and combining the different strategy iteration and reinforcement learning theory. The simulation results show that compared with the traditional controller, the safe reinforcement learning control method can effectively limit the exhaust temperature of the aeroengine within a safe range and optimize its control effect.