Predictive Control via Augmented Lagrange Function for Autonomous Vehicles Trajectory Tracking with Dynamic Quantization
Abstract
This paper investigates the Augmented Lagrange Function (ALM) based model predictive control (MPC) strategy for autonomous vehicle trajectory tracking. Firstly, a linear time-varying error model with measurement uncertainty and dynamic quantization is established. Measurement uncertainty indicates that vehicle state extraction fails due to the jittering of sensors during actual driving. The dynamic quantizer, which can adjust the quantization parameters online, is used to quantize the control input signals. The stability condition of the closed-loop system is obtained through the Lyapunov function and solved using linear matrix inequality technology. Then, according to the stability condition, the model predictive controller is designed by solving a “min-max” optimization problem that is based on a cost function over the finite time horizon. In order to solve the MPC optimization problem, the regularized least-squares method with ALM and an online iterative algorithm are explored, which obtain the analytical solution of the optimization problem. Finally, the effectiveness of the designed controller is verified by simulation experiments which show that the ALM has a very accurate computational value.
DOI: 10.61416/ceai.v26i2.8868