Rack and pinion drives (RPDs) are key components of battery-swapping systems (BSSs) used in electric heavy trucks; the faults occurring in these drives reduce the efficiency, accuracy, quality of battery swapping, and even pose potential safety risks. The operating conditions of RPD driving gear in BSSs are characterized by speed fluctuations, relatively low speeds, and reciprocating motion. To assess the driving gear fault characteristics under these conditions, based on the solution of image recognition under complex and low illumination conditions, this study proposes a fault diagnosis framework that includes adaptive down-sampling, three-dimensional acceleration data fusion, multi-scale local binary pattern (MS-LBP) extraction, and sparse representation. First, adaptive down-sampling is used to smooth out the speed fluctuation. Subsequently, MS-LBP extraction is employed to obtain efficient fault features at low speed. Finally, dictionary learning and sparse representations are conducted on the MS-LBP features. The effectiveness and advantages of the proposed diagnosis approach are demonstrated using monitoring data acquired from a BSS. Moreover, comparative studies demonstrate that the proposed fault diagnosis method yields improved performance.

As a critical component of aircraft flight control systems, aileron actuator’s fault directly affects the performance of flight control system and the overall flight safety of the aircraft. Therefore, effective fault diagnosis of aileron actuators becomes particularly important. However, with the development of redundancy design, the structure of aileron actuators is more and more complex, and the fault modes are more and more diverse, which increases the difficulty of fault diagnosis significantly. To address this challenge, this study proposes a fault diagnosis method based on parallel-symmetrical dot pattern (SDP) and polar sparse representation. First, the current residual signals of force motors are obtained using a bi-step observer. Then, the residual signals are transformed into parallel-SDP images, where each of the four petals of the parallel-SDP image is generated from the corresponding channel. The parallel-SDP image presents global fault information and captures subtle changes of residual signals, which increases the sensitivity of fault diagnosis. Subsequently, rapid single-channel fault localization is achieved by barycenter analysis of the parallel-SDP images. Finally, based on the result of fault localization, the proposed polar sparse representation algorithm is utilized for fault diagnosis of mechanical and control faults separately. Based on the dataset obtained from physical-parameter-simulation, the proposed method was validated, and the accuracy of fault diagnosis was 99.29%, which is better than conventional fault diagnosis, meanwhile, the computational resources consumption of the proposed method is much less than deep-learning-based method, which is suitable for embedded computing system in aircraft.

Pedestrian positioning multi-agent systems are an important application of inertial measurement unit (IMU), because they can operate independently of external facilities. To suppress pedestrian positioning errors based on zero velocity update assistance, multisensor collaborative technology provides compensation for errors generated by low-cost inertial sensors. However, traditional multisensor information fusion methods suffer from limited enhancement and insufficient adaptability. In this study, four IMUs are assigned to the foot, shank, thigh, and waist, respectively. According to the periodic characteristics of pedestrian gait, a lower limb kinematic model is constructed. Moreover, based on the kinematic model, the differential constraint relationship equations of the four position vectors of the lower limb are derived. In addition, the attitude information and zero velocity point information characteristics of the foot and waist were analyzed and compared. The optimal waist attitude information and foot zero velocity information are used as observations, and an improved extended Kalman particle filter algorithm based on the foot–waist vector difference constraint is proposed. By real-time correction of foot–waist observation information, a distribution function that is more similar to the real distribution is established. The effectiveness of the proposed method has been verified for some specific shape trajectories, including 2-D and 3-D trajectories. Experiments have shown that compared with classic single-sensor methods and multisensor collaborative methods, the proposed improved method reduces positioning errors by more than 45% and 34%, respectively.

As the core energy storage unit of fast-swappable battery cases in electric heavy trucks, power batteries are particularly prone to connection loosening faults under complex operating conditions, such as frequent battery swapping, continuous vibration, and mechanical shocks. These faults are highly concealed and difficult to detect at an early stage, posing significant challenges for fault detection and diagnosis. To address this issue, this study proposes a vibration-electric synergy-based method for fault detection, localization, and diagnosis. In this framework, fault detection is used to determine whether an abnormal condition occurs, fault localization is employed to identify the faulty parallel branch, and fault diagnosis aims to recognize the specific fault type. First, the dynamic coupling relationship between vibration and current signals is accurately captured through linear correlation analysis, enabling efficient fault detection. Subsequently, by comparing the effective values of parallel branch currents, precise fault localization is achieved. Finally, leveraging multi-scale and multi-resolution analysis, three-dimensional signal feature point cloud construction, and a three-dimensional convolutional neural network, accurate fault diagnosis is realized, achieving an overall accuracy of 98.33%. Comparative and ablation experiments demonstrate that the proposed method significantly outperforms traditional machine learning approaches as well as certain deep learning methods, confirming its effectiveness and superiority. Furthermore, this study introduces a vibration-electric synergy-based framework for analyzing connection loosening faults, elucidating the causal relationship in which structural vibration induces electrical response evolution through variations in contact states, thereby providing a theoretical foundation for cross-domain studies.

Accurate relay fault diagnosis in electric heavy trucks’ High-Voltage Power Distribution Systems (HV PDS) is critical for safety, yet challenged by the high hardware cost of multi-channel sensing and the trade-off between robustness and computational efficiency in existing methods. This paper proposes a simplified monitoring system using only bus current and voltage signals, introducing a novel diagnosis method based on Operating State Transition Rules and a Feature Space Topological Analysis Classifier (FSTAC). First, state transition rules logically localize actuating relays through multi-domain feature analysis without extra sensors. Second, a cross-frame feature fusion mechanism extracts complementary time-frequency features for three typical faults, which are classified by the low-cost FSTAC algorithm. Benchmarked against traditional thresholding and deep learning models, experiments demonstrate that our approach achieves 98.69% localization and 99.25% diagnosis accuracy with merely 0.07s inference time. These results confirm the method’s superior balance of precision, robustness, and real-time ease of deployment for onboard HV systems.