Optimizing look-ahead distance for the pure pursuit has been an important issue as it determines the relation between the smooth path tracking performance and tracking error. This paper presents a novel model-based look-ahead distance optimization method to ensure both smooth path tracking performance and acceptable tracking error for pure pursuit. The conventional pure pursuit method was modified to give the look-ahead distance as the control input, and the optimization problem was formulated as model predictive control to optimize the look-ahead distance. The proposed method optimized the look-ahead distance by reducing it only when the vehicle was about to deviate excessively from the path and maximizing it elsewhere to improve cutting corner problems and ensure smooth path tracking performance. Simulation and experimental results of a test vehicle performing on various roads, including handling course, double lane change, and 90-degree turn, showed the ability of the proposed method to optimize look-ahead distance, gathering the advantages of the pure pursuit method with changing look-ahead distance.

Using a special type of sliding mode controller, a new type of traction control system for hybrid four-wheel drive vehicles is developed. This paper makes two major contributions. First, a new electric powertrain architecture with an in-wheel motor at the front wheels and a clutch on the rear of the transmission is proposed for maximum traction force. The in-wheel motors are controlled to cycle near the optimal slip point. Based on the cycling patterns of the front wheels, the desired wheel speed for rear wheel is defined. The rear wheels are controlled to track this defined speed by controlling the clutch torque. Unlike conventional TCS algorithms, the proposed method exploits clutch control instead of brake control. Second, a special type of sliding mode controller which uses a nonlinear characteristic of the tire is proposed. An important distinction between the proposed sliding mode control method and other conventional feedback controllers is that the former does not depend on feedback error but provides the same functionality. Therefore, the practical aspects are emphasized in this paper. The developed method is confirmed in simulations, and the results reveal that the proposed method opens up opportunities for new types of traction control systems.

This paper introduces a new torque vectoring (TV) algorithm with regards to enhancing the cornering performance of the electronic-four-wheel drive (e-4WD) vehicles. The proposed TV algorithm aims to distribute the in-wheel motor (IWM) torques to the left and right wheels assisting the vehicle in following the driver's intended trajectories. The proposed TV algorithm first included a yaw rate controller reference for neutral-steering which leads to the improvement of vehicle cornering agility. Thereafter, in-vehicle sensors and a standalone global positioning system (GPS) combined with a smooth sliding mode controller (SMC) were used based on a vehicle bicycle model to generate desired yaw moment values to track the reference. Finally, a novel torque distribution algorithm using the Daisy-chaining allocation, which is suitable for redundant actuator configuration, was utilized. Noteworthy, the Daisy-chaining allocation algorithm takes into account a torque operation area with both the T-N curve of IWM and the tire friction circle. In this manner, based on the vehicle bicycle model and using the CARSIM interface, the simulations, evaluation and verifications of the proposed control technique were done. Thereafter, considering a real-car based experiments with various driving scenarios, the effectiveness of the proposed TV algorithm was evaluated. It was confirmed that some evaluation factors in terms of the cornering performance were improved. The following main contributions make the proposed TV algorithm be a meaningful solution to enhance the cornering performance of e-4WD vehicles: 1) improvement of both smoothness and convergence rate of control action, 2) a practical and intuitive way to distribute IWM torques, and 3) high applicability to mass-produced vehicles.

In this paper, we propose an autonomous emergency steering (AES) control strategy for the automated vehicle to avoid collisions with obstacles while maintaining the vehicle's yaw stability. Specifically, by capturing the surrounding traffic environment (i.e., dynamic/static obstacles and lane-road boundary) using an artificial potential field approach, a safe vehicle path trajectory in the form of a polynomial function was generated via Pontryagin's maximum principle (PMP). The generated path was complemented by a linear time-varying model predictive control (LTV-MPC) scheme that can predict the state behaviors near the operating points during the prediction horizon. Therefore, the proposed approach can effectively and dynamically adapt to the changes in the system model while respecting the state-and-control constraints. Furthermore, we assumed harsh driving environments where the vehicle drives on a slippery road at a relatively high speed, which may cause lateral instability to the vehicle. Hence, the developed controller has to consider both the vehicle's yaw stability and the path-tracking performance. To achieve this goal, we imposed the adaptive constraints in the designed LTV-MPC, depending on the road surface and driving conditions to prioritize the satisfaction of vehicle yaw stability instead of the path-tracking performance. The proposed approach was verified under various conditions using high-fidelity vehicle dynamics and control testing software (i.e., CarSim). Its effectiveness in vehicle yaw stability and path-tracking performance is compared to conventional baseline approaches.

In a typical traffic scenario, autonomous vehicles are required to share the road with other road participants, e.g., human driven vehicles, pedestrians, etc. To successfully navigate the traffic, an adaptive robust level-k framework can be used by the autonomous agents to categorize the agents based on their depth of strategic thought and act accordingly. However, mismatch between the vehicle dynamics and its predictions, and improper classification of the agents can lead to undesirable behavior or collision. Robust approaches can handle the uncertainties, however, might result in a conservative behavior of the autonomous vehicle. This paper proposes an adaptive robust decision making strategy for autonomous vehicles to handle model mismatches in the prediction model while utilizing the confidence of the driver behavior to obtain less conservative actions. The effectiveness of the proposed approach is validated for a lane change maneuver in a highway driving scenario.