The over-actuated characteristics of distributed drive electric vehicles (DDEVs) provide a flexible platform to pursue higher holistic performance. This paper proposes a dual-model predictive control (MPC)-based hierarchical framework to realize the energy saving while improving the handling stability for DDEVs. The upper layer allocates the torque vector through the front/rear axles, which can provide a high-efficiency zone for the in-wheel motors and reduce the energy consumption. The lower layer generates a direct-yaw-moment control input by differential longitudinal forces of the left/right wheels to ensure the vehicle handling stability. Considering the time-varying state variables, a linear-time-varying model-predictive-control (LTV-MPC) method is adopted to guarantee the accuracy of the model. The combined magic formula tire model is used to modify the tire parameters, including tire longitudinal stiffness and cornering stiffness. The soft constraint constructed by β-γ phase plane is introduced in the LTV-MPC to ensure the vehicle stability, based on which, a relaxation factor is designed to reduce the energy consumption due to the excessive direct-yaw-moment control inputs. The simulation and hardware-in-the-loop (HIL) test results show that the proposed control framework can effectively reduce the energy consumption for DDEVs while ensuring the vehicle handling stability.

The direct yaw moment control system of distributed drive electric vehicles provides a flexible control scheme to enhance the vehicle stability through four independently driven in-wheel motors. However, the vehicle nonlinearities may lead to the invalidation of control. To this end, a Takagi–Sugeno (T-S) fuzzy-based robust H∞ control method is proposed to ensure the vehicle performance while addressing the nonlinear challenge. First, thanks to the T-S fuzzy modeling technology, the tire nonlinear characteristics are described by fuzzy rules, based on which the vehicle lateral dynamics model is established. Next, the safety region represented by the tire slip angles phase plane is presented to evaluate the vehicle stability performance. Considering the control priority of vehicle handling performance and stability control with different stability margins, a multiobjective optimization function is transformed into a standard robust performance optimization problem with dynamic weight coefficients. A T-S fuzzy-based robust H∞ state feedback controller is then designed to ensure the system stability and H∞ performance. Finally, the hardware-in-the-loop tests are conducted to validate the proposed controller. Comparative results show the effectiveness to improve the vehicle handling performance while ensuring the stability.

The rapid development of the active front-wheel steering technique makes it possible to pursue precise, fast, and reliable path-tracking control for autonomous vehicles. To mitigate the impact of time-varying model parameters and system disturbances on the control performance, this paper proposes a continuous-time solving scheme for the autonomous path- tracking problem, which combines the robust model predictive control (RMPC) technique, non-parallel distributed compensation (non-PDC) method, and polytopic model. First, considering the time-varying longitudinal vehicle velocity, a polytopic model is utilized to construct a vehicle path-tracking system with a simpler structure. Then, a state-feedback controller for the continuous system model is developed based on the non- PDC method. By using the Lyapunov stability proof to transform the optimal control problem into the linear matrix inequalities (LMIs) solving form, an RMPC method is used to ensure path-tracking accuracy. Moreover, a robust compensation control is introduced to improve the anti-disturbance and transient response of the system. The virtual simulation by Carsim/ Simulink and field vehicle tests are conducted to assess the proposed controller. Both the simulation and field vehicle tests prove the effectiveness of the proposed controller in improving the path-tracking performance.

Human-machine mutual trust has become one of the important factors restricting steer-by-wire vehicles due to the removal of mechanical connections. To this end, this paper proposes a hierarchical shared steering control framework based on the human-machine mutual trust evaluation. The upper level of the framework aims to evaluate the human-machine mutual trust level. The driver's trust level in the machine is evaluated by the human-machine steering difference. And the machine's trust level in the driver is evaluated by driver skills. The lower level is to dynamically optimize the authority allocation considering varying human-machine mutual trust states. The fuzzy method is adopted to calculate the reference value of the human-machine authority based on the mutual trust level. Through minimizing the lateral acceleration and tracking error, the fuzzy rule database for the authority reference level is further tuned offline. Furthermore, to improve the smoothness of authority transfer and path tracking, the model prediction control is used to optimize the human-machine authority levels online. The proposed authority allocation strategy is verified with the driver-in-the-loop test bench. The results show the effectiveness of improving driving performance under different human-machine mutual trust levels.

The uncertainties of driver's behavior seriously affect road safety and bring significant challenges to the human-machine cooperative control. This paper proposes a human-machine shared control framework considering driver's time-varying characteristics to improve the co-driving cooperation performance. Firstly, the driving intention is introduced to describe the driver's involvement level through using Gauss-Bernoulli restricted Boltzmann machine method. And the index of driving ability is proposed to evaluate driver skills based on path-tracking errors. Then, a novel human-machine authority allocation strategy is designed by combining the two driving behavior characteristics and used to construct the driver-vehicle interaction system. Subsequently, a T-S fuzzy robust state-feedback shared control system is developed considering time-varying driver behaviors and vehicle states. Finally, the proposed shared steering system is validated by the driver-in-the-loop test bench. The results show that the proposed control method can reduce human-machine conflicts and has obvious superiority in improving performance of driving comfort, path tracking, and vehicle stability for the co-driving vehicles.

High-speed cruising scenarios with mixed traffic greatly challenge the road safety of autonomous vehicles (AVs). Unlike existing works that only look at fundamental modules in isolation, this work enhances AV safety in mixed-traffic high-speed cruising scenarios by proposing an integrated framework that synthesizes three fundamental modules, i.e., behavioral decision-making, path-planning, and motion-control modules. Considering that the integrated framework would increase the system complexity, a bootstrapped deep Q-Network (DQN) is employed to enhance the deep exploration of the reinforcement learning method and achieve adaptive decision making of AVs. Moreover, to make AV behavior understandable by surrounding HDVs to prevent unexpected operations caused by misinterpretations, we derive an inverse reinforcement learning (IRL) approach to learn the reward function of skilled drivers for the path planning of lane-changing maneuvers. Such a design enables AVs to achieve a human-like tradeoff between multi-performance requirements. Simulations demonstrate that the proposed integrated framework can guide AVs to take safe actions while guaranteeing high-speed cruising performance.