Applied nonlinear control of vehicle stability with control and state constraints. (January 2020)

• Record Type: Journal Article
• Authors: Tahouni, Amin; Mirzaei, Mehdi; Najjari, Behrouz
• Is Part Of: Proceedings of the Institution of Mechanical Engineers. Volume 234:Number 1(2020)
• Page Start: 191
• Page End: 211
• Publication Date: 2020-01
• Subjects: Vehicle stability -- nonlinear optimal control -- input and state constraints -- nonlinear optimization -- prediction; Mechanical engineering -- Congresses; Transportation engineering -- Congresses; 629.2
• DOI: 10.1177/0954407019848858
• Languages: English
• ISSNs: 0954-4070

Abstract

For the vehicle dynamic control system, to guarantee directional stability in risky maneuvers, the side-slip angle should be restricted to the admissible range when the yaw rate tracks the proposed desired response for enhanced steerability. Meanwhile, the control input of the external yaw moment produced by asymmetric braking forces should be calculated in the practical range according to the capacity of tire forces. In the present study, a novel constrained controller with input and state constraints is developed. To this aim, a cost function consisting of predicted continuous response of yaw rate tracking error is expanded in terms of current control signal. Concurrently, the state constraint of side slip is transformed to the equivalent constraint of control signal by a novel nonlinear prediction approach. After that, the expanded performance index is analytically minimized in the presence of all input constraints to obtain the control law. The computed yaw moment is optimally distributed to asymmetric braking forces by designing a wheel slip control system. Simulation studies are conducted to evaluate the performance of proposed constrained controller compared with the unconstrained controller and a conventional nonlinear model predictive controller developed in the recent papers using a 14-degree-of-freedom vehicle model which includes suspension system dynamics. The results show that the proposed controller is much faster and easy to solve and implement.