The driving range of electric vehicles (EVs) is still an important factor restricting their development. Although the rising battery energy density has reached a bottleneck, which is a key constraint, the drive energy management strategy also has a significant effect and can improve the driving range of EVs, since wheel traction torque control can directly optimize the driving energy consumption of EVs. In order to comprehensively analyze the current research status of driving energy management and clarify its development direction, this review focuses on the driving energy management strategy of EVs and systematically summarizes the configurations and power distribution strategies of the dual-motor coupling drive system (DCDS), as well as torque vectoring control strategies of the decentralized drive system. Firstly, driving energy losses are analyzed in detail, which mainly include electric loss, tire slip energy dissipation, and the power of cornering resistance. Secondly, typical configurations of the DCDS are introduced, and the power distribution strategies of the DCDS are comprehensively reviewed. Finally, as an interesting energy-saving technology, energy-saving torque vectoring, generally applied to decentralized drive systems, is reviewed in detail in terms of its energy-saving pathways and control strategies, which are classified as front-and-rear torque vectoring and left-and-right torque vectoring. Research findings indicate that the driving range of EVs can be effectively increased by applying a driving energy management strategy based on several novel multi-power source drive systems. The development of a driving energy management strategy and the required novel drive systems will be a valuable and crucial direction for further energy conservation in EVs.

This study proposes the use of a Quadruple Deep Q-Network (QDQN) for optimizing the energy management strategy of Plug-in Hybrid Electric Vehicles (PHEVs). The aim of this research is to improve energy utilization efficiency by employing reinforcement learning techniques, with a focus on reducing energy consumption while maintaining vehicle performance. The methods include training a QDQN model to learn optimal energy management policies based on vehicle operating conditions and comparing the results with those obtained from traditional dynamic programming (DP), Double Deep Q-Network (DDQN), and Deep Q-Network (DQN) approaches. The findings demonstrate that the QDQN-based strategy significantly improves energy utilization, achieving a maximum efficiency increase of 11% compared with DP. Additionally, this study highlights that alternating updates between two Q-networks in DDQN helps avoid local optima, further enhancing performance, especially when greedy strategies tend to fall into suboptimal choices. The conclusions suggest that QDQN is an effective and robust approach for optimizing energy management in PHEVs, offering superior energy efficiency over traditional reinforcement learning methods. This approach provides a promising direction for real-time energy optimization in hybrid and electric vehicles.

In order to revise the deviation caused by ignoring the dynamic character of fault propagation in traditional fault propagation path identification methods, a method based on the maximum occurrence probability is proposed to identify the key fault propagation path. Occurrence probability of fault propagation path is defined by dynamic importance, dynamic fault propagation probability and fault rate. Taking the fault information of CNC machine tools which subject to Weibull distribution as an example, this method has been proven to be reasonable through comparative analysis. Result shows that the key fault propagation path of CNC machine tools is not unique, but changes with time. Before 1000 hours, key fault propagation path is electrical component (E) to mechanical component (M); after 1000 hours, key fault propagation path is auxiliary component (A) to mechanical component (M). This change should be taken into account when developing maintenance strategies and conducting reliability analysis.

In order to model and evaluate the reliability of long-life high-voltage relays with small-sample fault data characteristics, a reliability analysis method integrating average rank, the minimum mean square distance empirical distribution function, and total least squares estimation is proposed. In the random truncation experiment, considering the influence of random truncation data, the average rank method is used to correct the rank of small-sample fault data; then, the optimal empirical distribution function for small-sample fault data is obtained through the minimum average square distance, which can overcome the impact of small-sample fault data randomness. Under the assumption of the Weibull distribution model, the total least squares estimation method is used for reliability model parameter estimations, and the linear correlation coefficient and d-test method are used for model hypothesis testing. If two or more distribution models pass the linear correlation coefficient test and the d-test simultaneously, the root mean square error and relative root mean square error are applied to determine the optimal reliability model. The effectiveness of this method is verified by comparing it with the maximum likelihood estimation method.

A tool remaining useful life prediction method based on a non-homogeneous Poisson process and Weibull proportional hazard model (WPHM) is proposed, taking into account the grinding repair of machine tools during operation. The intrinsic failure rate model is built according to the tool failure data. The WPHM is established by collecting vibration information during operation and introducing covariates to describe the failure rate of the tool operation. In combination with the tool grinding repair, the NHPP-WPHM under different repair times is established to describe the tool comprehensive failure rate. The failure threshold of the tool life is determined by the maximum availability, and the remaining tool life is predicted. Take the cylindrical turning tool of the CNC lathe as an example, the root mean square error, mean absolute error, mean absolute percentage error, and determination coefficient (R²) are used as indicators. The proposed method is compared with the actual remaining useful life and the remaining useful life prediction model based on the WPHM to verify the effectiveness of the model.