To enhance the handling stability and active safety of Steering-By-Wire (SBW) vehicles, this study proposes a stability control strategy based on sliding mode control, integrating yaw rate and sideslip angle as feedback variables on the foundation of variable steering ratio control. A Particle Swarm Optimization-Grey Wolf Optimization (PSO-GWO) algorithm is applied to optimize the variable steering ratio with constant yaw rate gain, ensuring high sensitivity at low speeds and stability at high speeds. In addition, an adaptive global fast terminal sliding mode controller is developed to suppress chattering, which generates an auxiliary front-wheel steering angle to further improve vehicle stability. The SBW system is modeled in MATLAB/Simulink and co-simulated with CarSim under double lane-change maneuvers on high-speed/high-friction and medium-speed/low-friction roads. Simulation results show that compared with the uncontrolled case and the variable steering ratio control strategy alone, the proposed control strategy effectively improves the vehicle dynamic response and stability. Under the high-speed/high-friction condition, the lateral displacement error is reduced from 6.37% to 3.57%; under the medium-speed/low-friction condition, it is reduced from 9.74% to 2.11%. This study enhances the stability and active safety of SBW vehicles and provides valuable insights for the development of advanced SBW systems.

To address the lateral stability control problem of four-wheel steer-by-wire vehicles, a control strategy based on Piecewise Affine (PWA) identification of tire cornering characteristics is proposed. The PWA identification method is employed to model the nonlinear tire cornering characteristics, based on which a Global Fast Terminal Sliding Mode (GFTSM) lateral stability control strategy is developed. The proposed strategy adopts a coordinated active steering architecture: the front-wheel controller aims to track the desired yaw rate and calculates the active front-wheel steering angle, while the rear-wheel controller considers the tracking errors of both yaw rate and sideslip angle and generates the active rear-wheel steering angle through a weighted fusion approach. Co-simulation results using MATLAB/Simulink and CarSim demonstrate that, under the low-adhesion double-lane-change condition, the proposed method reduces the maximum sideslip angle by 57.6%, 73.5%, and 98.2% compared with the PWA-based Linear Quadratic Regulator (LQR) control, linear tire model-based GFTSM control, and proportional control, respectively. Under the high-adhesion large-steering-angle condition, the peak sideslip angle is reduced from 4.852 3° to 0.065 9°, while the peak yaw rate is decreased from 19.53 (°)/s to 16.07 (°)/s. These results verify that the proposed method can effectively suppress vehicle lateral state fluctuations and improve the lateral stability of FSV under extreme operating conditions.

To address the difficulty in coordinating path tracking accuracy and lateral stability of multi-axle steering vehicles under complex conditions such as high-speed driving and low-road adhesion, this paper proposes a path-stable tracking control method that integrates fuzzy control with integral sliding mode control. The proposed method calculates the front-wheel steering angle based on a driver preview model to track the desired path, and obtains the desired yaw rate and desired sideslip angle through a linear two-degree-of-freedom vehicle model. In the controller design, the yaw rate error and sideslip angle error are weighted and combined into a composite control error, based on which an integral sliding mode controller is constructed. Active rear-axle steering is then used to coordinate path tracking performance and vehicle stability. A fuzzy controller is further introduced to adjust the weighting coefficient online according to vehicle state errors, thereby improving the adaptability of the controller under different driving conditions. Simulink-TruckSim co-simulation results show that, under the high-speed and high-adhesion condition, compared with the conventional sliding mode control method, the proposed method reduces the maximum, average, and root mean square path tracking errors by approximately 26.5%, 43.3%, and 28.3%, respectively. Under the medium-speed and low-adhesion condition, the average and root mean square errors are reduced by approximately 24.2% and 3.8%, respectively. A low-speed double-lane-change real-vehicle test further verifies the feasibility and path tracking effectiveness of the proposed method.

To address the degradation of object detection caused by reduced contrast and blurred textures in foggy images, this paper proposes Fog-RT-DETR based on the Real-Time Detection Transformer (RT-DETR). The method integrates an Asymptotic Feature Pyramid Network (AFPN) for multi-scale feature fusion, a Variance-Enhanced Convolutional Block Attention Module (VE-CBAM) for local low-texture perception, and a Feature Degradation Enhancement Module (FDEM) for spatially weighted feature enhancement. Experiments on a foggy-weather dataset constructed from DAWN and ACDC show that Fog-RT-DETR achieves 81.45% mAP@0.5 and 58.78% mAP@0.5∶0.95, improving over the baseline by 2.12 and 3.89 percentage points, respectively. Recall increases by 0.49 percentage points. Ablation experiments show that combining all three modules yields the best detection results in the reported experiments. Channel compression experiments further demonstrate a trade-off between detection accuracy and computational cost.

Regarding the problem of thermal runaway in automotive lithium-ion batteries, in this paper the experiments were performed based on 1 Ah soft pack batteries, Adiabatic Accelerating Rate Calorimetry (ARC) and lateral heating experiments. The thermal runaway behavior and changing process of silicon-based negative batteries with different SiO_x mass fractions (0%, 5%, 10%, 15%) was researched. It is studied that the influence of silicon content on the characteristic temperature of thermal runaway and the battery voltage, as well as the changing characteristics of flame jetting during thermal runaway. The quantitative relation between silicon content-SEI stability-thermal runaway was established. The results show that the thermal stability index (temperature threshold) T₁ of the SEI film gradually decreases with the increase of SiO_x, and the triggering temperature T₂ of thermal runaway increases, and the highest temperature T₃ gets to 30.3% (when the SiO_x content is 15%). At the same time, comparing with lithium-ion batteries made of graphite materials, the unit heat release increases by 42.3%. In the lateral heating experiment, the highest temperature of battery thermal runaway gets to 1 003.5 ℃ (when the SiO_x content is 15%), and the flame jetting is more intense and the duration is shorter (10 s).

In light of traditional reliability and durability testing standards have lagged behind, and where conventional user survey methods struggle to accurately reflect user needs, this paper integrates user big data with public road big data. Taking road surface roughness and driving speed as core reference points, it establishes a method for correlating user behaviour with durability testing conducted at test tracks. Specifically, this method first establishes a road surface roughness map of public roads. By linking and coupling users'connected vehicle driving data via latitude and longitude coordinates, the driving data for each user is discretised to construct a four-dimensional feature matrix comprising vehicle speed, road surface roughness, load mass and mileage. Subsequently, based on collected segments of typical road load spectra, the load spectra for each user are precisely reconstructed to obtain the target loads for each user; Finally, test track durability test specifications are formulated based on the damage equivalence principle. The study established a road surface roughness map of public roads in Liuzhou City. Based on load spectra from 16 typical roads covering a cumulative distance of over 1 000 km at various speeds, load spectra for 2 000 users totalling nearly 7×10⁶ km were constructed, and the corresponding test track durability test specifications were formulated accordingly.

To address the issues of premature convergence and insufficient parameter adaptability of the Rime Optimization Algorithm (RIME) when dealing with high-dimensional mixed variables, as well as the problem of the LSTM's reliance on manually set hyperparameters which affect the diagnostic performance, a fault diagnosis method based on the improved RIME algorithm optimizing LSTM is proposed and applied to the multi-condition fault identification of automotive gearbox rolling bearings. Firstly, this study makes three key improvements to the original RIME algorithm to overcome its premature convergence and insufficient parameter adaptability when dealing with high-dimensional mixed variables: introducing an adaptive parameter adjustment strategy based on population diversity to balance the exploration and exploitation capabilities of the algorithm; designing a Levy flight global disturbance mechanism to enhance the ability to escape from local optima; proposing a mixed search space encoding scheme to efficiently handle continuous and discrete hyperparameters. Simulation experiments on standard test functions show that the improved RIME algorithm has significantly improved convergence speed and solution accuracy. Secondly, on this basis, the improved RIME algorithm is applied to optimize the key hyperparameters of LSTM, and an improved RIME-LSTM fault diagnosis model is constructed. Finally, experimental research is conducted using the self-built comprehensive simulation test bench for automotive gearbox rolling bearing faults. The experimental results show that the proposed improved RIME-LSTM fault diagnosis method has a diagnostic accuracy rate of up to 99.98% and an F₁ score of 99.75%, which is significantly superior to the comparison methods such as PSO-LSTM and GWO-LSTM. At the same time, the noise robustness test shows that this method can still maintain the highest diagnostic performance under different noise levels.