To achieve accurate prediction of electric vehicle remaining range，a method based on a three-layer weighted stacking model for predicting remaining range of electric vehicles is proposed in this paper. By combining the maximal information coefficient and Spearman correlation coefficient as criteria for variable evaluation，the minimum redundancy maximum relevance algorithm is employed to optimize and obtain the input feature set from the candidate features. A three-layer stacking model that incorporates the original training features is then constructed，and Bayesian optimization algorithm is used to determine the weights of the base models within the stacking model. Finally，the input feature set is used to train the three-layer weighted stacking model and realize electric vehicle remaining range prediction. The results show that the proposed three-layer weighted stacking model has high prediction accuracy and，compared to other models，with stronger generalization capabilities.

High performance active suspension has significant advantages in improving driving experience，and robust control algorithm is an important guarantee for active suspension performance. To solve the problem that the typical robust control algorithms are difficult to achieve effective disturbance estimation and compensation，in this paper，a H₂/H_∞-H₂-IUDE algorithm is proposed to estimate and compensate the disturbance by using IUDE algorithm and introducing in H₂ state observer，which improves the robustness compared with H₂/H_∞ algorithm. Firstly，the model of half vehicle active suspension control systems is established，and the disturbance form is defined. Then，an IUDE algorithm for disturbance estimation and compensation decoupling is proposed，and a H₂ state observer is proposed to redesign H₂/H_∞ algorithm. Finally，simulation analysis is carried out for typical working conditions such as random road surface and speed bump road surface. The results show that，compared to the H₂/H_∞ algorithm，the proposed algorithm reduces the root mean square values of the vehicle body center vertical acceleration and pitch angle by 7.6% and 5.9%，respectively，under random road conditions，demonstrating a significant improvement in vehicle ride comfort. Meanwhile，the proposed H₂ observer can effectively estimate system states. The IUDE algorithm can accurately estimate disturbance，and can avoid the deterioration of suspension dynamic deflection caused by the non-decouple UDE method，which has outstanding characteristics of excellent disturbance estimation and flexible compensation.

Aluminum alloy integrated die-casting technology is an important means to achieve automotive lightweight. In this paper，structure optimization，molding simulation technology，die-casting technology are adopted for "material-structure-process" integrated design and manufacture of shock tower. Firstly，combined with extensive design experience，the three-dimensional data of the shock tower is determined through performance objectives，material selection and structural design. Subsequently，the performance analysis of the shock tower is conducted to ensure that the performance objectives are met，and the mold flow analysis of the temperature and velocity fields is conducted based on the designed mold structure，and optimization measures are proposed. Finally，the parts trial production is carried out to conclude problems and analyze causes，and propose solutions to form the shock tower development process. The study shows that the weight reduction rate of the shock tower reaches 16.5% while meeting the requirements of various performance indexes. The integrated design method is feasible，provides the industry with the ability to analyze the whole process and actual production experience，and enhances the confidence of manufacturers in adopting integrated die-casting technology.

In order to conduct more comprehensive safety analysis of electric cars in side pole collision scenarios，in the paper a locally refining scheme is used to build the finite element model of battery pack，which is applied to simulation at both the whole car and the battery pack levels. Based on accident statistics，the side pole collision responses of the car are examined by changing the impact positions，angles and collision speed，including the conditions defined in the national standards. Considering the high cost of the whole car simulation，a parameterized model of battery pack level is established by adjusting collision speed and mass compensation for large-scale side pole collision simulation. A fast prediction model based on the energy method is proposed for battery pack level collision safety，which can predict the deformation and mechanical failure risk in real time under different side pole impact conditions. The model is validated with an average prediction error of 3.22%.

To study the impact of undulating road on driving stability of hazardous chemical liquid tank truck，the medium and long wave road models are constructed based on road roughness data. Taking a multi-axles liquid tank truck as the object，a vehicle dynamics model considering the tire deformation and suspension nonlinear characteristics is established. Equivalent mechanical models of the longitudinal and lateral sloshing of the liquid are established，which are coupled with the vehicle dynamics model. Kinetic parameters of the tank on different roads are obtained to excite the forced liquid sloshing model. The results show that the increase in the phase difference between the two sides of the road has an adverse effect on the lateral stability of the truck. The frequency of liquid lateral sloshing decreases with the increase of road wavelength and increases with the rise of liquid filling ratio. When the distance between the tractor saddle and the axle of the semi-trailer is an integer multiple of the road wavelength，there is significant longitudinal liquid sloshing. The oblique wave deflector can suppress lateral liquid sloshing，but the suppression effect is poor when the liquid filling ratio is low under small sloshing conditions. When the truck passes through uneven roads with a low filling ratio，increasing the vehicle speed appropriately can reduce the liquid sloshing. While the filling ratio is high，excessive vehicle speed will lead to a significant increase in liquid sloshing amplitude and wall load，reducing the driving stability of the vehicle.

Energy management determines the power distribution of the power system of fuel cell vehicles （FCVs） and affects the economy and durability of FCVs. As the operating conditions of vehicles are complex and variable，energy management can improve the output performance of FCV power system by integrating traffic information. In this paper，the optimization objectives of FCV energy management are summarized，and the traditional rule-based and optimization-based energy management strategies are analyzed. Then，focusing on the analysis and prediction of traffic information such as vehicle speed and traffic condition，prediction methods such as Markov method and artificial intelligence are reviewed，and the research progress of FCV energy management strategies by integrating traffic information is summarized. Finally，the research direction of development of FCV energy management by integrating traffic information is proposed.

In order to develop a hood with better pedestrian protection performance，firstly three types of straw energy absorption modules are designed，specifically circular，triangular and rectangular. Subsequently，the CSEAM sample with PA11 material is prepared and the finite element model is verified. Then，the human body injury protection effectiveness of the three types of Novel Straw-shaped structure Sandwich Hood （NSSH） is evaluated by simulation and the deformation mode is analyzed. The results show that the three NSSHs can significantly reduce pedestrian head injury，and the Triangle Straw-shape structure Sandwich Hood （TSSH） has the best protective effectiveness on human body injury. The deformation is uniformly diffused in a regular shape，and the peak of acceleration and impact force in each stage of the collision waveform is optimized by reducing the overall structural stiffness of the hood. Further analysis shows that the human body injury protection effectiveness of TSSH decreases with the increase of vehicle speed and impact angle，and the injury caused by TSSH is significantly lower than that of the original hood，indicating that raising the hood can reduce the impact angle to better protect pedestrians，while when the Ratio of Pedestrian height to Bonnet leading edge height R_h≤2.1 and R_h≥2.45，the head only impacts the hood and windshield，respectively. It can be used as the criterion of hood lift. The research results can provide support for the design of hood structure with better pedestrian protection effectiveness.

Megacasting during rapid filling and cooling processes inevitably generates defects，which significantly impact the mechanical properties of castings. However，it is difficult for existing mechanical analysis models to accurately predict the properties in defective castings，posing substantial challenges to the design of structurally safe castings. To solve the problem，a constitutive model and a fracture criterion of cast aluminum considering defects are proposed in this paper. Five different shapes of samples are cut from different locations of the megacasting floor，and experimental tests are carried out. The defect information on the fracture sections is statistically analyzed using scanning electron microscopy. Based on the stress-strain curves of the standard tensile samples，a constitutive model considering defects and saturating stresses is proposed to accurately characterize the strain-hardening properties. Based on the existing Modified Mohr-Coulomb （MMC） fracture criterion，an improved MMC model considering defects and stress states is proposed，and the parameters are calibrated by four different shaped samples. To validate the effectiveness of the proposed model，a comparative analysis between experimentation and simulation is conducted. The results show that the proposed constitutive model has a high fitting accuracy compared with the traditional hardening model. The load-displacement curves of the tests and simulation of different samples are in good agreement，which verifies the validity of the proposed fracture model. This study provides a novel approach to accurately predicting the mechanical properties of megacasting aluminum alloys.

For the dynamic multi-objective requirements of yaw maneuverability and lateral stability under different driving states，an adaptive control strategy for vehicle yaw stability is proposed. Firstly，the vehicle dynamics model is established by piecewise linear fitting technology，and the dynamic stability region boundary related to road adhesion and longitudinal velocity is obtained in phase plane by integrated application of improved two-line method and fuzzy theory. Then，the stability risk during vehicle driving is quantitatively characterized and the dynamic multi-objective mapping function is introduced to adjust the built-in parameters of the stability control strategy designed based on the model predictive theory so as to match the dynamic multi-objective demand of vehicle lateral stability，yaw maneuverability and actuator energy consumption. Finally，the simulation test results prove that the designed control strategy can help the vehicle obtain safer and better stability control effect than the traditional method under various conditions.

For the problem that it is difficult for video vehicle detection models to extract rich target features in complex traffic monitoring scenarios，in this paper a new spatial-temporal feature fusion module SF-Module is established from the perspective of making full use of spatial-temporal feature information of video images. The multi-head self-attention mechanism in Transformer model is used to extract and fuse the temporal and spatial feature information of current and historical frames of video vehicle images to enrich the feature information of the target. On this basis，based on YOLOv8 network，the newly created spatio-temporal feature fusion module SF-Module is integrated in its neck network to mine spatio-temporal feature information of video image sequences. At the same time，the WIoU loss function is introduced as the prediction frame regression loss to reduce the harmful gradient generated by the low quality label frame，and the SFW-YOLOv8 video vehicle detection model is designed. Finally，the newly established SFW-YOLOv8 complex scene video vehicle detection model is tested on the UA-DETRAC dataset，and some images in the dataset are simulated to enhance the data on rainy and foggy days，so as to improve the generalization of the vehicle detection model. The experimental results show that the values of mAP50 and mAP50：5：95 of the SFW-YOLOv8 video vehicle detection model are 79.1% and 63.6%，which are 1.7% and 3.3% higher than that of the YOLOv8 model，respectively. The reasoning speed is 11 ms/ frame，which has excellent detection performance.