Automated Design Synthesis of LQR Controllers for Road Vehicle Active Suspension Systems

摘要

This paper presents an automated design method for road vehicle active suspension controllers. In order to improve ride quality of road vehicles, controllers based on the Linear Quadratic Regulator (LQR) technique have been explored for active suspension systems of road vehicles. In conventional designs of LQR controllers for road vehicle active suspension systems, the weighting factors of the corresponding performance index are determined using trial and error method; the governing equations of motion for vehicle models are derived by hand. This is a tedious and time consuming process. To address this problem, an automated design method is proposed: the vehicle system model is generated using a multibody system software package, a genetic algorithm is utilized to search optimal weighting factors for the LQR controller, and the design optimization is implemented via parallel computing. To demonstrate the effectiveness of the proposed design method, it is applied to the LQR controller design for active suspensions of a road vehicle represented by a quarter-vehicle model. The simulation results show that the resultant LQR controller derived by the proposed method outperforms its counterpart based on conventional trial and error method in terms of displacement and vertical acceleration of sprung mass and suspension deflection. Moreover, the parallel computing technique significantly reduces numerical optimization time.