For ordinary drivers, stabilizing drift during collision avoidance is challenging, potentially leading to serious accidents. To enhance driving safety and human-machine cooperative efficiency under such extreme conditions, this paper presents a human-centered drift-curbing (HCDC) control strategy considering the driver’s acceptance of the driving assistance system’s intervention. Specifically, the acceptance domain (AD), a framework for capturing and describing the driver’s intervention tolerance characteristics, is presented. To constrain the additional steering intervention within the driver’s AD, a soft-intervention (SI) strategy based on steering ratio control is formulated. On this basis, an HCDC controller is developed using nonlinear model predictive control (NMPC), with individualized constraints designed by referring to the driver’s AD. Driver- and controller-in-the-loop verification experiments are conducted in a low adhesion highway collision avoidance scenario prone to drift. The experimental results indicate that the proposed HCDC control strategy provides an acceptable “soft” intervention, enabling drivers to stabilize drift quickly while reducing steering effort.

In formulating a driver-oriented human-machine shared control (HMSC) system, an intelligent driving system (IDS) should ideally minimize interventions under normal driving conditions while providing timely and necessary interventions in hazardous situations. To achieve this balance, this paper proposes a novel driver-oriented active intervention control (DAIC) framework for automotive collision avoidance (CA). As a foundation, a spatiotemporal collision risk evaluation (SCRE) method is developed by integrating collision probability (CP) and time-to-collision (TTC) metrics, comprehensively quantifying collision risks in both spatial and temporal dimensions. Subsequently, a short-horizon driver behavior prediction network (SDBP-Net) based on bidirectional long short-term memory (BiLSTM) is proposed, enabling effective prediction of short-horizon driver input increment sequence and corresponding vehicle trajectories. Leveraging these developments, a partially cooperative game-based HMSC strategy is formulated to dynamically allocate control authority between the driver and IDS. Driver-in-the-loop (DIL) experiments validate the effectiveness of the proposed DAIC framework. The results demonstrate that the DAIC framework significantly reduces unnecessary IDS interventions while maintaining collision safety, effectively balancing driving safety and driver acceptance. This work highlights the practical potential of the feedforward HMSC system in achieving safer, more intuitive and driver-friendly collision avoidance assistance.

As heavy vehicles advance toward increased intelligence and modernization, the control of advanced driver assistance systems for ensuring driving safety faces significant challenges. To enhance the driving safety of heavy vehicles operated by drivers with varying driving styles, this article proposes a human–machine cooperative control (HMCC) strategy that combines steering and braking using deep deterministic policy gradient (DDPG) algorithm. First, a multiagent system is adopted as the framework for the driving safety assistance control system, wherein the active front steering (AFS) system and the differential braking control system (DBC) function as subsystems. These subsystems interact through control sequence information while managing yaw and roll stability. The optimal control performance of both the AFS and DBC is ensured using a distributed model predictive controller and Pareto optimality theory. Second, to analyze different drivers’ driving styles, safety characteristic parameters were collected from multiple drivers. By analyzing the effects of drivers on yaw and roll stability, drivers were classified into three types. Furthermore, an HMCC strategy based on DDPG is designed. Phase plane constraints that consider yaw and roll stability are incorporated into the design of the DDPG reward function, training the agents to allocate cooperative control weights between the driver and the AFS and DBC controllers. Finally, the proposed control strategy’s effectiveness is validated through the electro-hydraulic compound steering and braking hardware-in-the-loop test system, demonstrating its ability to improve driving safety for different driver characteristics.

To enhance the trajectory tracking capability of steer-by-wire (SBW) vehicles while reducing driver’s workload, a human-machine shared control (HMSC) strategy using improved reinforcement learning-based model predictive control (RL-MPC) is proposed. In this paper, two main contributions have been made: 1) for driver loop, a variable steering ratio (VSR) strategy applied to SBW system is designed based on an improved fuzzy controller, whose parameters are optimized through simulated annealing (SA) algorithm; 2) for intelligent control system loop, an improved RL-MPC method is proposed to realize the high precision steering tracking control for autonomous vehicles (AVs), in which MPC and deep deterministic policy gradient (DDPG) are deeply integrated to combine their short-term optimization ability and long-term value estimation capability. Moreover, to shorten the time of overall training and ensure that the optimal control strategy can be explored, the DDPG agent is pretrained before the parallel training of RL-MPC. CarSim-MATLAB/Simulink co-simulation results show that in the whole tracking process, the lateral position error and yaw angle error of the vehicle are significantly reduced, indicating that the tracking accuracy is greatly improved. Meanwhile, the steering wheel angle and speed are also reduced, which means that the driver will spend less energy during the steering process.

To improve the overall performance of human-vehicle cooperation and enhance the drivers’ confidence in the advanced driver assistance system (ADAS), an adaptive haptic assistance control scheme for the steer-by-wire (SBW) vehicle is presented in this paper. A comprehensive human-vehicle system model is built, including vehicle dynamics, the SBW model, and the driver’s arm neuromuscular dynamics model, as a foundation for controller design. An expert driver model based on a multi-layer feed-forward neural network (MLFN) is developed to generate the reference steering angle for haptic assistance design. The individual driver’s arm characteristics are identified and incorporated into the adaptive haptic assistance controller design to generate personalized torque assistance, facilitating a typical driver to achieve the same trajectory-tracking performance as experts. The nonsingular fast terminal sliding mode (NFTSM) is applied to calculate the assistance torque to ensure the fast finite-time convergence and robustness of the system. Simulations and driver-in-the-loop experiments are conducted, with results showing that the proposed haptic assistance controller can help drivers complete the trajectory-tracking task by providing personalized torque assistance while reducing their steering workload.

The road information preceding the vehicle serves as a fundamental prerequisite for intelligent driving systems. In comparison to other data requirements of autonomous vehicles, road information exhibits a more two-dimensional character. Among the array of sensors currently equipped on autonomous vehicles, it is only the visual sensors that possess the capability to detect and interpret the road surface. The majority of road recognition algorithms rely heavily on multi-classification methods, which, however, tend to introduce significant data imbalance issues. The process of further refining pavement types exacerbates this problem. Based on the Road Surface Classification Dataset (RSCD), this study uses the diffusion model to generate category images with less data, and uses the traditional image augmentation method to augment the original image dataset. Furthermore, a convolutional neural network-based multi-label classification model, incorporating the Zero-bounded Log-sum-exp& Pairwise Rank-based loss functions, is trained to enhance the model's predictive performance and accuracy. This enhanced model ultimately achieved an average accuracy of 86.8%, demonstrating significant improvements in performance.

In order to explore the nature of the human-vehicle game problem and analyze the torque-angle interaction between the two agents, i.e. the driver and advanced driver assistance system (ADAS), a human-vehicle game stability control strategy based on Nash negotiation principle is proposed. First a six-order vehicle dynamic model, a driver neuromuscular (NMS) model, etc., are set up to simulate drivers’ steering characteristics and vehicles’ response to the input of the two agents and external disturbance. Secondly, several significant parameters in NMS model are recognized, and an active rear steering (ARS) controller is designed using the sliding mode variable structure algorithm. Then, the Nash negotiation solution is worked out according to Nash negotiation principle, and a self-tuning method for the weight of ARS controller is put forward employing the fuzzy control theory. Finally simulations are carried out under the standard double-lane change maneuver. The results indicate that the stability control strategy proposed in this paper can effectively solve the game problem and achieve good vehicular stability control performance.

This paper proposes a fault-tolerant control method for distributed-drive electric vehicles (DD-EV) considering individual driver steering characteristics that may vary among human drivers. Both vehicle and driver models are explicitly utilized for the fault-tolerant control design with considerations on modeling inaccuracies. A sliding-mode control scheme is generated to tolerate the DD-EV actuator fault and control the vehicle motions in a tailored cooperation with the specific driver. Such a personalized and fault-tolerant control method can specifically assist the human driver in post-fault vehicle motion control and reduce both physical and mental workloads of the driver. Co-simulation results of the controller using Matlab and CarSlm®indicate that the control law can provide appropriate control assistance to different drivers, achieving effective human-vehicle control cooperation in post-fault vehicle maneuvers.