A Dual Closed-loop Longitudinal Speed Controller Using Smooth Feedforward and Fuzzy Logic for Autonomous Driving Vehicles

The performance of longitudinal control directly affects the safety and ride comfort of autonomous driving vehicles. For autonomous driving vehicles, the driving torque depends on the engines characteristics, which is difficult to build accurate model. In addition, the nonlinear and response lag of the power system will lead to drastic changes and oscillations of speed tracking, which will significantly reduce the longitudinal control performance of vehicle speed. Therefore, in order to secure the safe longitudinal control of a vehicle, we propose a dual closed-loop longitudinal speed controller using smooth feedforward and fuzzy logic, that is, smooth acceleration feedforward is introduced in the speed-loop to reduce the response lag, and fuzzy rules are formulated in the acceleration-loop to adaptively adjust the parameters of PID controller. This scheme has been implemented on the autonomous driving platform 'Pioneer'. The experimental results on the real car show that, compared with conventional methods, the proposed longitudinal control scheme with smooth feedforward and fuzzy logic can obtain higher speed tracking accuracy and gentler control characteristics of throttle/brake pedal actions.