Parameter Self-tuning of ADRC for Electronic Throttle Based on CARLA Algorithm

The electronic throttle system (ETS) is a critical component of the automotive gasoline engine, and ensuring accurate and rapid tracking of the disired throttle position without overshoot poses significant importance for vehicle safety and driving comfort. In this study, we designed an active disturbance rejection controller (ADRC) to achieve the accurate throttle tracking control. Considerring performance degradation caused by the engine aging and changes in operating conditions, a continuous action reinforcement learning automata (CARLA) algorithm is applied to tuning the control parameters adaptively. Simulation results demonstrate the effectiveness of the proposed parameter tuning strategy.