To address the issue of insufficient estimation accuracy of the Hydraulic Control Unit (HCU) wheel cylinder pressure estimation algorithm for an integrated wire controlled hydraulic braking system with eight valve structure, this article proposed wheel cylinder hydraulic pressure estimation algorithm that can replace pressure sensors. The article first proposed a wheel cylinder hydraulic pressure estimation algorithm based on the Bernoulli principle, and then analyzed the basic characteristics of the hydraulic adjustment unit and the Pressure Volume (PV) characteristics of the brake fluid based on hardware in the loop experimental bench testing. Finally, it was verified through actual vehicle experiments. The experimental results show that under basic braking conditions, the root mean square error of the proposed wheel cylinder pressure estimation algorithm is within 0.259 MPa, the root mean square error under active braking conditions is within 0.374 MPa, which is equivalent to the accuracy of the scheme with added pressure sensors.

In order to study the seat safety of passenger cars in China, this paper analyzed the casualties of each seat in three collision modes based on China In-Depth Accident Study (CIDAS) (2011~2022) statistical data, calculated the fatal risk of passengers in the front and rear seats using the risk model, and used risk weighting model and geometric average model of the risk to calculate risk of each seat. The results show that the fatal risk coefficient of the front row is 1.18 compared with the back row. Taking the risk of 100% of the driver’s seat position as the reference standard, the risk of the passenger seat, the left rear seat, the right rear seat and the middle rear seat was 79.57%, 105.23%, 93.28% and 191.69%, respectively.

To predict injury of the occupant in vehicle collision more rapidly and accurately, a training database for deep learning models was established based on frontal 100% overlap rigid barrier real-world collision data, and data preprocessing and features extraction were conducted. Deep learning models were constructed separately based on Long Short-Term Memory (LSTM), Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) neural network, and Temporal Convolutional Networks (TCN) for injury prediction training. The validation results show that the model prediction accuracy reaches 0.8579, 0.8209 and 0.9674, respectively, demonstrating feasibility of the proposed method.

In order to enhance the safety of the battery system in electric vehicles under side pole impact, a finite element model incorporating the battery system was established. The accuracy of this finite element model was validated through vehicle side pole impact tests. Using this model, the deformation status and acceleration response of the battery pack in side rigid pole impact scenarios were analyzed with structural deformation and acceleration impact as the primary evaluation criteria. The results indicate that under side pole impact, both the external framework and internal support structure of the battery pack experience significant deformation, and all positions within the battery system withstand substantial acceleration impacts. Based on this result, the paper proposed safety-specific design for this impact condition when designing power battery pack and vehicle body sill beam structure.

In order to improve the efficiency of reliability verification of the e-Axle under mechanical load and thermal load environment, this paper, taking the three-in-one e-Axle as the research object, analyzed the failure mechanism and influencing factors of each subsystem by comprehensively considering the actual operating environment and working conditions of the e-Axle, and studied the acceleration process and calculated equivalent test time according to the accelerated failure logic under different test conditions. The results show that the accelerated life endurance test method can support the reliability verification of the e-Axle.

In order to investigate the effect of frontal collision deceleration waveform parameters on dummy chest injury, the method of equating the deceleration waveform to a Bi-Slope Approximation Wave (BSAW) was used to complete the frontal collision simulation using MADYMO. The results show that the dummy damage corresponding to the bilinear deceleration waveform curve with a larger slope in the first stage and a higher peak at the end is smaller under the condition that the vehicle speed and deformation are guaranteed to be constant; the influence of BSAW parameters on the chest deceleration peak along the X-direction is from the first stage peak, the second stage peak, and the first stage peak moment in decreasing order under the condition that speed change remains unchanged.

Based on the experimental method and technical requirements of GB 15086—2013, a finite element analysis model was established to analyze inertial force resistance and optimize plastic tailgate structure. Firstly, the plastic tailgate structure failure motion mode and stress of inertial force resistance was predicted and analyzed based on the structural design feasibility of plastic tailgate by using finite element analysis model simulation; secondly, the issue of structural failure and cracking of plastic tailgate under inertial load were solved by adopting the method of structural optimization scheme through increasing the load transmission path of the tailgate inner plate and strengthening the local material thickness. Finally, the reliability of the optimization scheme of the plastic tailgate and the accuracy of the simulation analysis model were verified by the bench test.

In the state-of-charge estimation of power battery, the traditional Extended Kalman Filter (EKF) ignores high-order terms and Particle Filter (PF) suffers from particle degradation and loss of diversity during the resampling process. To address this issue, this paper proposed the improved Mixed Kalman Particle Filter (MKPF) algorithm. Firstly, the extended Kalman filter was used to generate the state estimate of the system, and then the unscented Kalman filter was used to repeat the process. The state estimates obtained by the extended Kalman filter and the unscented Kalman filter were used together as the particle filter proposal distribution, and value sorting was used to determine the survival of the fittest particles. Simulation and experimental results show that the maximum error of SOC estimate by the proposed algorithm is 1.2%, which is better than the estimation accuracy of the existing PF, EKF, and UKF algorithms on SOC.

With a light truck equipped with power-following fuel cell as the research object, a power-following energy management strategy of a fuel cell based on wavelet transformation has been constructed combining the frequency characteristics of the vehicle power demand under typical driving cycles using wavelet analysis method. The low frequency component of the vehicle power demand obtained after wavelet transform was allocated to the fuel cell, so as to improve the fluctuation of the fuel cell output power. In addition, the wavelet decomposition level was dynamically adjusted to regulate the power following degree automatically, to ensure the power supply of the vehicle under various power battery charge-discharge capacity and vehicle power requirement. The simulation result shows that the power-following allocation strategy based on wavelet transform can meet the vehicle power demand under typical driving conditions, and the SOC can be kept in the reasonable range, the power output fluctuation of the fuel cell has been decreased by 29.8%, in addition, the hydrogen consumption has been reduced by 1.5%.

In order to solve the problem of low accuracy in predicting the remaining service life of proton exchange membrane fuel cells, this paper proposed a dynamic fuel cell Remaining Useful Life (RUL) prediction model based on Northern Goshawk Optimization (NGO), Convolutional Neural Network (CNN) and Bi-directional Long Short-Term Memory (BiLSTM) neutral network. Firstly, NGO optimized the learning rate, hidden nodes and regularization coefficient of the CNN-BiLSTM model, and then the CNN-BiLSTM model extracted the features of the input data through the convolutional layer, and input it into the BiLSTM layer for timing modeling and prediction. In addition, wavelet threshold de-noising algorithm was used to smoothen the original data. Pearson correlation coefficient was used to extract model input variables, and NGO-CNN-BiLSTM network power prediction model was built. The simulation and verification results show that this method can effectively improve the prediction accuracy of the remaining service life of fuel cells up to 99.49%, which is higher than that of other comparative models.