Power batteries are the core of electric vehicles, but minor faults can easily cause accidents; therefore, fault diagnosis of the batteries is very important. In order to improve the practicality of battery fault diagnosis methods, a fault diagnosis method for lithium-ion batteries in electric vehicles based on multi-method fusion of big data is proposed. Firstly, the anomalies are removed and early fault analysis is performed by t-distribution random neighborhood embedding (t-Sne) and wavelet transform denoising. Then, different features of the vehicle that have a large influence on the battery fault are identified by factor analysis, and the faulty features are extracted by a two-way long and short-term memory network method with convolutional neural network. Finally a self-learning Bayesian network is used to diagnose the battery fault. The results show that the method can improve the accuracy of fault diagnosis by about 12% when verified with data from different vehicles, and after comparing with other methods, the method not only has higher fault diagnosis accuracy, but also reduces the response time of fault diagnosis, and shows superiority compared to graded faults, which is more in line with the practical application of engineering.

The application of electric motor emulators (EME) has been prosperous in recent years along with the advancement in power hardware-in-the-loop technology. However, EME design methods based on a typical modeling approach and control structure tend to be invalid in promising a clear, dynamic accuracy objective and a robust dynamic accuracy performance over wide working points, especially at high operating speed. Hence, a new EME design method based on multi-input multioutput (MIMO) modeling and a MIMO closed-loop current control (MCLCC) structure is proposed in this article. Singular values of the MIMO motor model are analyzed to clarify the influence of the rotor speed on the dynamic accuracy objective of EME for the first time, based on which an unknown input observer is designed in the MCLCC structure to compensate for disturbance voltages caused by modeling errors. The transient tracking errors of both current and torque responses can thus be remarkably reduced at 300-Hz electrical speed in the experiment even if 35% modeling errors exist, verifying the superiority of the proposed EME design method in dynamic accuracy.

This article proposes a novel space vector pulsewidth modulation (SVPWM) scheme for multilevel converters in the abc coordinate system, in which the reference vector is decomposed into an offset vector and a remainder vector that can be synthesized using two-level SVPWM. The switching state satisfying a specific relationship is first selected as the offset vector such that the common-mode components are eliminated, and therefore, the phase voltages are decoupled. A dynamic reference point mechanism establishes a one-to-one correspondence between all available vector combinations and one unique variable, i.e., the number of level shifts. This feature further facilitates a general approach to determining the optimum switching states to satisfy different control objectives, such as common-mode voltage rejection. Besides, phase decoupling enables the introduction of carrier-based modulation to simplify implementation, where the duty cycle of the zero vector can be flexibly adjusted, while the nonzero vectors remain the same to generate an equivalent output. Consequently, the two-level SVPWM-based method and carrier-based pulsewidth modulation are combined to exploit their respective strengths. The proposed scheme achieves overmodulation operation and provides more flexibility, i.e., redundant switching states and adjustable duty cycles, to optimize switching patterns. Simulation and experimental results validate the proposed algorithm.

The electric machine emulator (EME), using digital simulation and power electronics to emulate the characteristics of actual machines, can greatly accelerate the testing of electric drives. However, most existing EMEs are based on typical L filter and linear controller, which causes control conflicts and bandwidth limitation. To address this issue, this paper presents an EME based on LCL filter with passive damping for a three-phase permanent magnet synchronous motor. To improve the dynamic emulating accuracy, a dual closed-loop deadbeat predictive current control algorithm is proposed, which is computationally efficient and easy to implement. The system stability is analyzed in the discrete domain, and the parameter constraints of the filter are obtained. Then, two unknown input observers are designed to compensate for the disturbance currents and voltages caused by modeling errors. Moreover, instead of the empirical method, a theoretical one considering the harmonic suppression, bandwidth, stability and resonance is presented for filter design. Finally, the performance of the proposed EME is validated through simulation and experimental results under various conditions such as machine start-up, torque step change, and speed reversal.

This article proposes novel space vector pulse width modulation strategies for multilevel converters to achieve two widely desired optimization objectives, common-mode voltage (CMV) elimination and switching frequency minimization (SFM). For CMV elimination, a new coordinate system that looks like the cropped version of the original abc frame is proposed, by which zero CMV can be ensured when the identified switching states are transformed from the new frame back into the original one. For SFM, a new approach is proposed to determine the optimal switching sequence, where a voltage reference point dwell mechanism is introduced such that the number of switching actions between adjacent switching periods is minimized by establishing its relationship with a unique variable, namely the number of level shifts. Furthermore, these two objectives can either be implemented independently or in combination to achieve SFM under the premise of CMV elimination, thus allowing the converter's redundancy to be fully exploited. Simulation and experimental results validate the proposed strategies.

Object detection is becoming increasingly critical in autonomous driving. However, the accuracy and effectiveness of object detectors are often constrained by the obscuration of object features and details in adverse weather conditions. Therefore, this paper presented the DAN-YOLO vehicle object detector specifically designed for driving conditions in adverse weather. Building on the YOLOv7-Tiny network, SPP was replaced with SPPF, resulting in the SPPFCSPC structure, which enhances processing speed. The concept of Hybrid Dilated Convolution (HDC) was also introduced to improve the SPPFCSPC and ELAN-T structures, expanding the network’s receptive field (RF) while maintaining a lightweight design. Furthermore, an efficient multi-scale attention (EMA) mechanism was introduced to enhance the effectiveness of feature fusion. Finally, the Wise-IoUv1 loss function was employed as a replacement for CIoU to enhance the localization accuracy of the bounding box (bbox) and the convergence speed of the model. With an input size of 640 × 640, the DAN-YOLO algorithm proposed in this study achieved an increase in mAP0.5 values of 3.4% and 6.3% compared to the YOLOv7-Tiny algorithm in the BDD100K and DAWN benchmark tests, respectively, while achieving real-time detection (142.86 FPS). When compared with other state-of-the-art detectors, it reports better trade-off in terms of detection accuracy and speed under adverse driving conditions, indicating the suitability for autonomous driving applications.