Seibum B. Choi

Kyoungseok Han, Giseo Park, Gokul S. Sankar et al.

This paper proposes a new control strategy to improve vehicle cornering performance in a model predictive control framework. The most distinguishing feature of the proposed method is that the natural handling characteristics of the production vehicle is exploited to reduce the complexity of the conventional control methods. For safety s sake, most production vehicles are built to exhibit an understeer handling characteristics to some extent. By monitoring how much the vehicle is biased into the understeer state, the controller attempts to adjust this amount in a way that improves the vehicle cornering performance. With this particular strategy, an innovative controller can be designed without road friction information, which complicates the conventional control methods. In addition, unlike the conventional controllers, the reference yaw rate that is highly dependent on road friction need not be defined due to the proposed control structure. The optimal control problem is formulated in a model predictive control framework to handle the constraints efficiently, and simulations in various test scenarios illustrate the effectiveness of the proposed approach.

Seungtaek Kim, Jonghyup Lee, Kyoungseok Han et al.

To address the computational challenges of Model Predictive Control (MPC), recent research has studied using imitation learning to approximate MPC with a computationally efficient Deep Neural Network (DNN). However, this introduces a common issue in learning-based control, the simulation-to-reality (sim-to-real) gap. Inspired by Robust Tube MPC, this study proposes a new control framework that addresses this issue from a control perspective. The framework ensures the DNN operates in the same environment as the source domain, addressing the sim-to-real gap with great data collection efficiency. Moreover, an input refinement governor is introduced to address the DNN's inability to adapt to variations in model parameters, enabling the system to satisfy MPC constraints more robustly under parameter-changing conditions. The proposed framework was validated through two case studies: cart-pole control and vehicle collision avoidance control, which analyzed the principles of the proposed framework in detail and demonstrated its application to a vehicle control case.