Design and Evaluation of EV Drivetrain Clunk and Shuffle Management Control System

Vehicle drivetrains contain backlash and compliance, making it challenging to transmit torque to the wheels smoothly and rapidly. Due to the presence of this lash and compliance, any rapid changes in the actuator torque gives rise to the so-called clunk and shuffle phenomena, which degrade the drivability of the vehicle. Hence, the actuator torque has to be shaped so that clunk and shuffle are maintained at acceptable levels.In order to accomplish the shaping, we develop a switching control system that switches between the contact mode and the backlash mode, based on an estimation of the drivetrain state. The system includes a controller, which consists of a Butterworth precompensator and a lead feedback compensator in the contact mode and a bang-bang controller in the backlash mode. It also includes an estimator, which is designed as an Extended Kalman Filter. We apply the proposed approach in an EV drivetrain, and show using simulations that the control system is effective in delivering the driver demand quickly, while, at the same time, also mitigating the clunk and shuffle.Along the way, we introduce metrics that quantify the performance of the shaping system, from the point of view of smoothness, responsiveness, and intensity of clunk. Based on these metrics, we show that the performance of the proposed system compares favorably with that of a torque shaping strategy, which is applied in some of our current vehicles.