TS-MPC for Autonomous Vehicles including a dynamic TS-MHE-UIO
For autonomous vehicle control, this work offers a computationally efficient method to handle nonlinear dynamics, though it is an incremental improvement over existing MPC techniques.
This paper presents a cascade control approach for autonomous vehicle trajectory tracking, combining a novel TS-MPC for position control and a TS-LMI-LQR for dynamic control, along with a TS-MHE-UIO estimator. The TS formulation reduces computation times by 40-50 times compared to nonlinear MPC.
In this work, a novel approach is presented to solve the problem of tracking trajectories in autonomous vehicles. This approach is based on the use of a cascade control where the external loop solves the position control using a novel Takagi Sugeno - Model Predictive Control (TS-MPC) approach and the internal loop is in charge of the dynamic control of the vehicle using a Takagi Sugeno - Linear Quadratic Regulator technique designed via Linear Matrix Inequalities (TS-LMI-LQR). Both techniques use a TS representation of the kinematic and dynamic models of the vehicle. In addition, a novel Takagi Sugeno estimator - Moving Horizon Estimator - Unknown Input Observer (TS-MHE-UIO) is presented. This method estimates the dynamic states of the vehicle optimally as well as the force of friction acting on the vehicle that is used to reduce the control efforts. The innovative contribution of the TS-MPC and TS-MHE-UIO techniques is that using the TS model formulation of the vehicle allows us to solve the nonlinear problem as if it were linear, reducing computation times by 40-50 times. To demonstrate the potential of the TS-MPC we propose a comparison between three methods of solving the kinematic control problem: using the non-linear MPC formulation (NL-MPC), using TS-MPC without updating the prediction model and using updated TS-MPC with the references of the planner.