This paper proposes a fuzzy PID control strategy with start-stop function for the direct battery cooling systems of electric vehicle. A whole vehicle model is developed using AMESim and Simulink co-simulation, to simulate the battery heat dissipation process under different control strategies. The results indicate that compared with traditional PID control, the start-stop fuzzy PID control has advantages including faster response time, shorter overshoot duration and lower system power consumption. Furthermore, the effect of shortened overshoot duration and power consumption savings is more pronounced in scenarios with higher ambient temperatures or lower driving speeds.

To address the problem that distributed-drive electric vehicles cannot accurately track the desired trajectory after steering system failure, this paper proposes a fault-tolerant control strategy that considers vehicle stability and trajectory tracking after steering failure. Firstly, the desired front wheel angle and additional transverse moment are calculated based on model prediction control and sliding mode control, respectively. Then, for the steering system failure fault, a front wheel angle tracking controller is designed based on non-singular terminal sliding mode control to solve the differential steering moment required to achieve the desired angle. Secondly, with the optimization objectives of minimizing the tire loading rate and the control volume error, the tire force distribution is realized based on the quadratic programming algorithm. Finally, simulation tests are carried out under two working conditions of medium-speed high adhesion coefficient and high-speed low adhesion coefficient respectively, and the results show that the proposed fault-tolerant control strategy can still make the vehicle track the desired trajectory stably after the steering system fails, and it has good control effect.

To solve the problem of unevenness and large difference in temperature at the air outlet of car air conditioners, temperature control curve of a car air conditioning outlet is simulated and tested. The correlation analysis of the outer length, inner length ard inclination angle of the deflector structure with the awerage temperature and temperature difference is performed by using response surface optimization to derive the regression polynomials, response surface graphs, and the optimal parameter design of the deflector structure. The simulation results show that this method improves the temperature control curve performance of the car air conditioner

In order to accurately estimate the battery parameters, state of charge and power state at different temperatures, a recursive least squares method combined with adaptive extended Kalman filter algorithm based on adaptive forgetting factor is proposed. By correcting and updating parameters in real time, the accuracy of battery parameter identification and state-of-charge estimation is improved. Based on the constraints of the model terminal voltage identification results, the state-of-charge estimation results and the maximum discharge current of the battery, the joint estimation of the battery power state is realized. The test results show that the maximum absolute error of the identification voltage and the maximum absolute error of the state of charge are 62.699 mV and 1.894%, respectively under the dynamic stress test condition. When the continuous discharge time is 5 s, 30 s and 120 s, the average error of battery power is 5.6×10^-3 W, 6.5×10^-3 W and 8.0×10^-3 W, respectively. The proposed adaptive joint estimation algorithm can improve the accuracy of parameter identification and state estimation effectively.

In order to address the issue of human-machine conflict caused by the neglect of human-machine interaction in traditional lane keeping assistance systems, this paper proposes a lane keeping human-machine co-driving strategy. A lane departure decision model considering the drivers’ lateral driving habits is designed by characterizing the drivers’ lateral driving habits based on their historical lateral positions and dynamically dividing the road boundary according to 3σ principles. At the same time, the control of the assisted driving system is allocated based on the risk assessment value and driver fatigue factor. The experimental results show that the proposed human-machine co-driving strategy can effectively avoid lane departure risks caused by fatigue driving and driving errors. The lane keeping assistance systems, which considers the lateral driving habits of drivers can provide drivers with sufficient freedom while applying appropriate constraints to suppress lane departure, effectively reducing human-machine conflicts and ensuring safety.

To address the issue of exterior rear view mirror whistle of a home-made passenger car, vehicle subjective evaluation is conducted to confirm and analyze the issue, then schemes are proposed using Computational Fluid Dynamics (CFD) to eliminate the unreasonable flush and cavity of exterior rear view mirror shell and camera, and improve stream field pressure of the monitoring point, vehicle test and subjective evaluation prove effectiveness of this scheme. The results show that exterior rear view mirror whistle problem can be solved by reducing Error State Index (ESI) which comprehensively considers factors such as road type, speed range and severity of NVH from 1.12 to 0.

In order to effectively extract interaction features among vehicles in high-speed traffic scenarios, thus accurately predict the trajectories of dynamic obstacles, this paper proposes a multi-vehicle interaction trajectory prediction model using the coding-decoding framework based on the graph spatial-temporal attention mechanism. The vehicle-to-vehicle graph interaction field is established by combining the repulsive force field and the graph model, the node feature matrix and the adjacency feature matrix are used to characterize the dynamic interaction between the vehicle and the surrounding vehicles, and the deep spatial-temporal interaction features are extracted by the graph spatial attention and temporal polytope attention to obtain the graph spatial-temporal fusion coding features. The one-hot encoding of the longitudinal and lateral behavior intentions of the vehicles is concatenated with the encoding to achieve multimodal trajectory prediction for the target vehicles. Validation using the NGSIM dataset shows that, compared with 6 other models, the proposed model achieves the lowest RMSE and NLL values. Ablation experiments further validate the effectiveness of the graph interaction field, demonstrating that the model can significantly improve the accuracy of vehicle trajectory prediction.

In order to enhance accuracy of driving fatigue detection, this paper takes drivers’ physiological signal pulse wave as the data source, introduces hemodynamic-based blood pressure waveform features based on the extraction of Heart Rate Variability (HRV) features. Moreover, a feature indicator set that can effectively characterize driving fatigue is constructed, and a three-classification model of driver fatigue is constructed based on ensemble learning. Then, a resampling method is introduced in the data preprocessing stage, and the effects of different sampling methods on the detection performance of the model are contrasted. Test results show that multidimensional feature fusion of pulse signals can significantly improve the detection accuracy of driver fatigue by 24.68 percentage points on average in all scenarios compared with the method of using only HRV features; resampling can further enhance the detection performance of the ensemble learning model, and the model achieves the best detection performance in a scenario with a sampling window width of 2 min, a sampling window overlap of 80%, and a fusion of HRV features with pulse waveform features.

With regard to the driving environment and safety of sanitation vehicle drivers, this paper proposes a driver fatigue detection method based on an enhanced YOLOv8n algorithm. Specifically, FasterNet is employed to replace the backbone network of the YOLOv8 object detection algorithm, resulting in the design of a lightweight FasterNet-YOLO network model. To preserve critical feature information from the input feature map, Squeeze-and-Excitation (SE) modules are integrated into the backbone network, while Convolutional Block Attention Modules (CBAM) are added to the neck network. Additionally, the Zero-DCE++ algorithm is introduced to enhance the brightness of video streams captured by cameras, addressing the issue of insufficient brightness in the driver’s face that hinders accurate detection. Experimental results demonstrate that the proposed method achieves an average precision of 98% (mAP@0.5) at an intersection over union ratio of 0.5, with an average inference time per frame reduced to 6.95 ms. This approach can effectively monitor the driver’s fatigue state in real-time under varying lighting conditions.

To investigate motion sickness caused by mismatched visual and operational information in the integration of autonomous driving and virtual reality technologies, this paper simultaneously collects electroencephalogram (EEG) signals from participants using a dual-task paradigm that combines active driving and autonomous driving on a simulated driving platform. This approach is complemented by the Go/No-go behavioral paradigm and standardized motion sickness questionnaires to explore the impact of different driving modes on the allocation of brain cognitive resources. Results indicate that autonomous driving scenarios significantly exacerbate motion sickness symptoms due to visual-vestibular conflict. Autonomous driving based on virtual reality is particularly prone to inducing motion sickness. The underlying neural mechanisms are characterized by increased power spectral density in the Pz, Cz, and Fz EEG channels (p<0.05), as well as decreased amplitude and shortened latency of the N200 and P300 components (p<0.05). Furthermore, a convolutional neural network classification model is constructed that integrates time-domain ERP, frequency-domain PSD, and nonlinear complexity features. The model achieves an accuracy of 92.7%, which provides a scientific basis for real-time monitoring and the optimization of human-computer interaction design.