In order to realize the optimization of autonomous navigation of mobile robot under the condition of partial environmental knowledge known. An improved Q-learning reinforcement learning algorithm based on prior knowledge is proposed to solve the problem of slow convergence and low learning efficiency in mobile robot path planning. Prior knowledge is used to initialize the Q-value, so as to guide the agent to move toward the target direction with a greater probability from the early stage of the algorithm, eliminating a large number of invalid iterations. The greedy factor ε is dynamically adjusted based on the number of times the agent successfully reaches the target position, so as to better balance exploration and exploitation and accelerate convergence. Simulation results show that the improved Q-learning algorithm has a faster convergence rate and higher learning efficiency than the traditional algorithm. The improved algorithm has practical significance for improving the efficiency of autonomous navigation of mobile robots.

This paper proposes a control strategy for the variable boost characteristics of electronic hydraulic brake (EHB) systems based on the driving style in response to the nonlinear challenges faced by the EHB systems in intelligent driving under complex personalized requirements. Initially, the working principle of the active braking of the EHB-booster was analyzed, and equivalent dynamic models and Karnopp friction models were established. Subsequently, by identifying the displacement and velocity parameters of the brake pedal, three types of variable boost characteristics-sporty, comfortable, and standard-were designed to satisfy the requirements of different driving styles. Then, to address the nonlinear disturbances caused by the variable boost characteristics, a variable-gain multiclosed-loop control strategy that considers nonlinear friction and inertia compensation was developed. Finally, the proposed control strategy was tuned and verified through the AMESim and Simulink cosimulation platform and vehicle tests. The results demonstrate that the strategy exhibits excellent control performance under various braking conditions that match driving styles, with steady-state control errors within 0.1 Mpa, providing a feasible solution for the complex nonlinear problems faced by personalized implementation in higher-order intelligent driving.

With the increasing demand for perception in complex road environments in intelligent driving, rapid and accurate identification of paved-road obstacles has become a critical prerequisite for driving safety and comfort. Various types of road obstacles can significantly affect vehicle stability and ride quality. To address this challenge, a lightweight and efficient vision-based obstacle detection framework, termed GroupScale-YOLO, is proposed, in which detection accuracy and computational efficiency are jointly enhanced through the collaborative design of multiple novel modules. First, a dedicated dataset targeting common paved-road obstacles is constructed, and six data augmentation strategies are employed to mitigate the adverse effects of road surface undulations and illumination variations on visual perception. Second, to overcome the limitations of YOLOv11n in paved-road obstacle detection tasks, targeted optimizations are introduced to the backbone network, convolutional blocks, and detection head. Experimental results indicate that GroupScale-YOLO achieves a 29.95% reduction in model parameters while simultaneously increasing mAP@0.5 by 0.6% on the self-built dataset, demonstrating its suitability for deployment in resource-constrained scenarios. Furthermore, real-vehicle road tests confirm that the proposed method maintains stable and accurate obstacle detection performance under practical driving conditions, offering a reliable solution for intelligent vehicle environmental perception.

In this study, to address the problems of the redundant safety and mass production of electro-mechanical braking (EMB) structures that are widely used in distributed drive electric vehicles (DDEV), we designed a compact dual-disc electromagnetic–EMB composite brake. The composite brake embeds an electromagnetic brake into the original friction disc, which realizes an organic combination of the friction and electromagnetic brakes. Electromagnetic braking has the advantages of no friction, a rapid response, and a high-speed braking effect, which can effectively improve the reliability and mechanical redundancy of composite braking systems. The braking system comprises regenerative, electromagnetic, and friction braking, which are typical hybrid systems. We designed a mode-switching control strategy for a composite braking system based on the hybrid control theory. MATLAB/Simulink were used to model each system and set different simulation conditions. The simulation results showed that, under different working conditions, the hybrid automata control strategy had a fast response speed, small overshoot error, and adapted to different road conditions. The feasibility of the redundant design of the electromagnetic–friction–regenerative composite braking structure and the rationality of the hybrid automata control strategy design were verified.

Diaphragm springs, as critical components in commercial vehicle clutch assemblies, directly determine the clutch’s working performance. The design of diaphragm springs, which possess a distinct symmetrical structure that underpins their mechanical behavior, centers on obtaining the large-end nonlinear load–displacement curve—a typical large deformation-induced nonlinear problem. Traditional design relies on the A-L formula, but studies show finite element analysis (FEA) yields results closer to actual measurements. This study established an FEA model of the diaphragm spring’s disc spring (excluding separation fingers) and validated its correctness by comparing it with the A-L formula. Then, using FEA on models with separation fingers, it analyzed factors influencing the large-end load–displacement characteristics. Leveraging the inherent symmetry of the diaphragm spring structure, particularly the symmetrical distribution of separation fingers, the analysis process efficiently captures uniform mechanical responses during deformation, while this symmetric arrangement also ensures balanced load distribution during clutch operation, a critical factor for stabilizing the load–displacement curve. Results indicate the separation finger root is a key factor, with larger root holes, square holes (compared to circular ones), and more separation fingers reducing stiffness to effectively adjust the curve; in contrast, the tip and length of separation fingers have little impact, making the latter unsuitable for design adjustments.

Previous trajectory tracking control methods do not take into account the continuity and correlation of the driving process, which may result in a low similarity with the real driver's motion state and manipulation behaviour. In this paper, a trajectory tracking control method based on driving segment extraction and splicing is proposed. A key characteristic of this method lies in the frame-by-frame extraction of driving manipulation data, vehicle trajectory, and motion state, which are then employed as the reference values for the preview-tracking controller to perform tracking tasks. Firstly, a framework for trajectory tracking control based on driving segment extraction and splicing is proposed, and the modelling and control process is standardized in a formal language. Secondly, a virtual driving platform based on PreScan+Simulink+CarSim is constructed to build the field, collect driving data from 30 testers, and build the driving segment database. Then, in order to reflect the wide range of validation results, driving type analyses are performed for all driving testers. Then, the trajectory tracking control algorithms applied to the driving segment extraction process and the splicing process are presented, respectively. Finally, simulation experiments and real vehicle experiments are designed to compare the Pearson similarity and feature parameters of manipulation behaviour and vehicle states of different drivers and different driving models. The results show that the new method can better simulate the driver's manipulation behaviour.