Accurate short-horizon trajectory prediction is crucial for safe and reliable autonomous driving. However, existing vision language models (VLMs) often fail to accurately understand driving scenes and generate trustworthy trajectories. To address this challenge, this study introduces KEPT, a knowledge-enhanced VLM framework that predicts ego trajectories directly from consecutive front-view driving frames. KEPT integrates a temporal frequency–spatial fusion (TFSF) video encoder, which is trained via self-supervised learning with hard-negative mining, with a k-means & HNSW retrieval-augmented generation (RAG) pipeline. Retrieved prior knowledge is added into chain-of-thought (CoT) prompts with explicit planning constraints, while a triple-stage fine-tuning paradigm aligns the VLM backbone to enhance spatial perception and trajectory prediction capabilities. Evaluated on nuScenes dataset, KEPT achieves the best open-loop performance compared with baseline methods. Ablation studies on fine-tuning stages, Top-K value of RAG, different retrieval strategies, vision encoders, and VLM backbones are conducted to demonstrate the effectiveness of KEPT. These results indicate that KEPT offers a promising, data-efficient way toward trustworthy trajectory prediction in autonomous driving.

The direct blending of polyether ether ketone (PEEK) with a solid lubricant such as polytetrafluoroethylene (PTFE) improves its tribological performance, but compromises its outstanding mechanical properties and processability. While these negative effects might be circumvented via the hybrid wear method, the influence of the contact temperature between multiple sliding components acting together is not fully understood. Herein, an analytical temperature model considering the influence of both micro- and macro-thermal behavior is extended to predict the contact temperature of a dual-pin-on-disk hybrid wear system. The interactions between several heat sources are investigated and experimentally verified. The analytical results show that the nominal temperature rise of the shared wear track is determined by the combined effect of the heat generated by both pin components, while the rise in flash temperature at the region in contact with each pin component is dependent upon its individual characteristics and working conditions. Hence, while different temperature peaks can coexist in the shared wear track, the maximum value dominates the performance of the system. For the experimentally investigated PEEK–PTFE–steel hybrid wear system, the formation of tribofilms is blocked, and the hybrid wear system fails, when the peak temperature exceeds the glass transition temperature of both pins due to an increase in applied load.

In order to improve the driving dynamics and riding comfort of pure electric vehicles, taking a two-speed I-AMT (Inverse-Automatic Mechanical Transmission) with rear friction clutch as the research object, a gear shift strategy, which consists of the open-loop control of the clutch position control and the closed-loop control of the drive motor speed control, is proposed. Considering the inherent time-delay and external disturbances within the motor speed adjustment system, a two DOF (degree-of-freedom) Smith predictor with feedforward input is designed to track the target speed of the drive motor. The feedforward input is used to eliminate the influence of clutch sliding friction on the motor speed control, while the feedback speed tracking controller is applied to realize the speed tracking performance with the existence of time-delay and the external disturbance. In order to verify the effectiveness of the gear shift control strategy and the accuracy of the two DOF Smith controller with feedforward control, simulation results comparison is firstly carried out to illustrate the effectiveness of the control scheme. Then, a light pure electric vehicle equipped with I-AMT was used for downshift experiments under large throttle, which is the most difficult working scenario to control the transmission. The experimental results show that the two DOF Smith controller can eliminate the influence of time-delay on the closed-loop control, and the proposed whole gear shift control strategy can limit the clutch slippage time within 1.5 s, resulting in a smaller shift jerk, thus guarantee the driving dynamics and riding comfort simultaneously.