Sensitivity analysis on weights of LQR for integrated chassis control

This article presents a sensitivity analysis on the weight of the linear quadratic regulator (LQR) for integrated chassis control. Electronic stability control (ESC) and active front steering (AFS) are used as actuators needed to generate a control tire force. After computing the control yaw moment in the upper-level controller, the lower-level controller was used to distribute the control yaw moment into the control tire forces generated by ESC and AFS. In this paper, the LQR was used to design the upper-level controller. The Taguchi method was used to complete a sensitivity analysis on the weight of the linear quadratic (LQ) objective function in the LQR. For the sensitivity analysis, simulation was performed using CarSim, a vehicle simulation package. From the sensitivity analysis, the effect of the LQ objective function weight on the control performance was checked. Based on this sensitivity analysis, it was determined that the weights on lateral acceleration and yaw rate error had the largest effect on the control performance.