State of charge (SOC) feature selection (FS) can significantly reduce model complexity and guarantee SOC estimation accuracy. However, there are limited FS methods since the electric bus driving dataset is a regression dataset with numerous continuous features. To tackle this issue, this paper proposes a novel FS method called dynamic maximum relevancy change and minimum redundancy change (DMRCmRC). Firstly, the modified strong approximation Markov blanket is constructed to obtain the feature groups and the count of strong relevant features in the regression dataset. Secondly, taking into account the interaction between the features in the dataset, three-dimensional fuzzy information measures are deduced to define the continuous candidate features’ dynamic relevancy change and redundancy change. Then, a novel feature evaluation function is designed to select the most critical features from each feature group, which employs a weighting scheme to combine feature relevance, redundancy, dynamic relevancy change and redundancy change. Furthermore, DMRCmRC has been tested on the STP, DEE, and BAS datasets. The tests show that DMRCmRC successfully improves the regression performance of feature subsets while reducing the feature dimension. Finally, DMRCmRC is applied to an electric bus driving dataset in Beijing, the statistical tests confirmed that DMRCmRC has significant statistical differences.

With the rapid development of electric vehicles (EVs) and their charging facilities, the coupling between the power distribution network (PDN) and the transportation network (TN) has been deepening. The increasing demand for fast charging of EVs can impact the stable operation of the coupled power and transportation network (CPTN). Therefore, accurate prediction of EV fast charging loads is crucial. In light of the limitations of traditional charging load prediction models that fail to fully consider traffic flow characteristics and user behavior, this paper proposes an EV fast charging load prediction model in CPTN considering users’ bounded rationality. A macro traffic flow model is employed to capture congestion dynamics within the TN, enabling a more realistic forecasting of traffic flow evolution. Additionally, a charging station dynamics model is developed to simulate the charging and queuing behaviors at charging stations, yielding the spatiotemporal distribution of charging loads. The micro user behavior decision model accounts for simultaneous route and departure time choices, reflecting the non-cooperative game behavior among user groups, and quantifies users’ bounded rationality as their tolerance for travel costs, leading to a more accurate equilibrium flow. By considering users’ bounded rationality, the model significantly reduces the iteration count, enhancing both computational efficiency and solution quality. Simulation results validate the effectiveness of the proposed model and algorithm, underscoring the necessity of accounting for traffic congestion characteristics and users’ bounded rationality.

Accurate analysis and prediction of low-temperature energy consumption in pure electric vehicles can provide a reliable reference for energy optimization strategies, thereby alleviating range anxiety. Here, we propose a data-driven energy consumption analysis and prediction approach for real-world electric vehicles in cold conditions. Specifically, the dataset was divided into multiple kinematic segments by the fixed-step intercept method, and principal component analysis was applied on segment parameters, showing the average speed and acceleration time had the greatest impact on energy consumption at −7 °C. Then, a Bayesian optimized XGBoost model, with the two factors above as input, was constructed to predict the cumulative driving and total energy consumption. This method was validated with two different types of pure electric vehicles under different dynamic driving cycles. The results demonstrated that the model could predict low-temperature energy consumption accurately, with all mean relative errors less than 3%.

An efficient deep reinforcement learning (DRL) energy management strategy for parallel hybrid electric vehicles is proposed in this paper. Firstly, a comprehensive model of the vehicle's powertrain system is established, and the energy management problem is briefly described. Subsequently, an efficient learning framework based on the deep Q-network (DQN) algorithm is constructed. The framework additionally incorporates a thermostat-based rule-aiding system to direct the model's training, accelerate the rate of training, and enhance the model's optimality and training rate in conjunction with enhanced priority experience replay, avoiding cold-start in RL. In the simulation, the proposed model is compared with methods based on DQN and rule-DQN. The results demonstrate that the newly developed DRL framework achieves higher training efficiency and optimality. Furthermore, this research examines how changes in the initial values of rule adoption rate and exploration rate influence the control performance of the model. Finally, the model, trained using a 520-second driving cycle, is evaluated on the worldwide light-duty test cycle (WLTC) driving cycle, validating its high adaptability.

Accurate state of charge and capacity estimation is crucial for battery packs in electric vehicles. However, the cell inconsistencies, computational complexity, temperature variations, and complex drive cycles all pose great challenges for the state of charge and capacity estimation of battery packs in field operation. This work aims to develop a framework for state of charge and capacity estimation of battery packs in real electric vehicles with few representative cells and physics-informed machine learning. By analyzing the battery data using a quantity- capacity diagram, the representative cells within the pack can be identified based on the relaxation terminal voltage online. Leveraging the physics information of batteries, the state of charge of the representative cells are monitored in a closed-loop manner. Then, state of charge and available capacity of battery pack are achieved based on the state of only two representative cells. We validate our method by evaluating its performance on the datasets of battery pack in real electric vehicles across different temperatures and drive cycles. The improvements of 83.60 % and 85.13 % in average root mean square errors can be achieved at real drive cycles under 30 degrees C and 15 degrees C. The results demonstrate that accurate and robust state estimation of battery packs can be achieved in field operations by utilizing the battery data of only two cells in laboratory conditions.

Electricity theft users with zero electricity usage (UZEU) should be specifically concerned in electricity theft detection (ETD) research. The challenges are: they provide no effective information on electricity usage behaviors, and they are easily confused with vacant house users. This has caused the majority of the existing detection methods relying on single electricity usage to fail to identify UZEU accurately. Hence, this article first analyzes the underlying correlation between water and electricity (W&E) usage collected by the smart meter. This analysis then lends the theoretical basis to propose a new ETD method by comprehensively using the multisource information. More precisely, the proposed method utilizes the mutual information coefficient (MIC) to construct a correlation model between W&E usage and in turn the wavelet clustering algorithm to cluster the MIC of the power distribution users. Thereafter, the resulting weak correlations indicate the suspected users as the electricity theft UZEU in case of zero electricity usage. Finally, the proposed method is validated by numerical experiments in the real world and illustrated to be more accurate than existing methods in detecting UZEU.

Accurate aging diagnosis is crucial for the health and safety management of lithium-ion batteries in electric vehicles. Despite significant advancements achieved by data-driven methods, diagnosis accuracy remains constrained by the high costs of check-up tests and the scarcity of labeled data. This paper presents a framework utilizing self-supervised machine learning to harness the potential of unlabeled data for diagnosing battery aging in electric vehicles during field operations. We validate our method using battery degradation datasets collected over more than two years from twenty real-world electric vehicles. Our analysis comprehensively addresses cell inconsistencies, physical interpretations, and charging uncertainties in real-world applications. This is achieved through self-supervised feature extraction using random short charging sequences in the main peak of incremental capacity curves. By leveraging inexpensive unlabeled data in a self-supervised approach, our method demonstrates improvements in average root mean square errors of 74.54% and 60.50% in the best and worst cases, respectively, compared to the supervised benchmark. This work underscores the potential of employing low-cost unlabeled data with self-supervised machine learning for effective battery health and safety management in real-world scenarios.

Thanks to the outstanding adaptability and relatively simple design process, machine learning-based energy management strategies (EMSs) show their superiority in reducing the energy consumption for multi-power powertrains. Given the popularity of distributed-drive in battery electric vehicles (BEVs), this study selects a dual-motor four-speed electric bus to investigate the proposed soft actor-critic (SAC)-based EMS. Two improvements are made to the traditional SAC algorithm to meet the special requirements of EMS in a distributed-drive electric bus in this study. Firstly, the introduced combination of Gumbel-SoftMax and actor-network allows the SAC agents to explore the hybrid action space (discrete operating modes and continuous torque distribution coefficients). Secondly, a heuristic rule-interposing action controller (HRIAC) is involved to reduce illogical exploration of SAC agents in searching for optimal power distributing ratio. Simulation results demonstrate that Gumbel-SoftMax facilitates agents' exploration and narrows the energy consumption gap between the proposed EMS and the global optimal EMS, meanwhile, HRIAC accelerates the convergence of agent training. The comparative results of proposed EMS performance in unknown driving cycles show that the proposed EMS outperforms other deep reinforcement learning (DRL)-based EMSs in adaptability, which is taken as a solid foundation for energy efficiency improvement of dual-motor electric bus in practice.

The booming development of the car-sharing industry and the continuous increase in the market share of electric vehicles demand effective dispatching strategies that consider collaborative optimization of route selection and charging management. Furthermore, in recent years, network attacks have posed significant threats to traffic management centers. Therefore, this paper presents a pioneering work aimed at investigating the vulnerability of electric vehicle sharing services in traffic management systems under network attacks. Firstly, we develop a bilevel optimization model to capture the game between network attackers and electric vehicle fleet operators. The upper-level problem represents the behavior of network attackers, who aim to maximize the total cost by manipulating traffic control signals and sensor data. The lower-level problem determines the system operator's decisions on optimizing route selection and charging scheduling to minimize costs. Secondly, a nested column generation algorithm is employed to solve the bilevel optimization model. Finally, several sensitivity studies are conducted using real-world traffic networks to examine the effectiveness of the proposed model and the influence of various parameters on system performance.

Distribution of lightning plays a key role in lightning protection of transmission lines. The design of wind deviation under lightning overvoltage of the transmission tower is an empirical parameter in the transmission line design code. Meteorological data from Hong Kong are investigated to analyze the joint distribution of lightning and wind speed. It has been uncovered that daily cloud-to-ground (CG) flashes follow the Burr distribution, which is highly skewed toward a few days with notable lightning. The lightning and wind follow a Gumbel copula joint distribution. According to empirical and theoretical distribution, there are 239 days with more than 1,000 CG lightning flashes per day, and approximately 20% of these days have a maximum wind speed of approximately or over 15 m/s. In 5 days with the number of CG lightning flashes above 30,000, 3 days have a maximum wind speed of over 15 m/s and the other 2 days have a maximum wind speed of over 10 m/s, which suggests that the severe convection with a squall line contributes much to the likelihood of the days with high wind and lightning storms.

With the ongoing proliferation of electric vehicle (EV) sharing systems, the need for dependable planning and operation of internet data centers and charging stations has become increasingly crucial. The article suggests a combined planning model that integrates a linked transportation system, power distribution system, and communication system (CS). The model combines the optimization of capacity and location for shared EV charging stations, Internet data centers, substations, and network expansion plans for both the power distribution and CSs. A new approach is presented for forecasting EV traffic flow by utilizing a multifactor based graph convolutional neural network and long short-term memory model that takes into account spatial-temporal correlations and external factors that may influence the flow. An enhanced weighting technique is employed to assess user preference factors, which are then integrated with projected EV traffic flow to anticipate shared EV demand. A practical experiment is conducted using simulations and data from an actual traffic scenario to assess the validity and efficacy of the proposed integrated planning model.

In the grid patrol work, there are some fault types with small targets in the line that needs to be detected. For the problem of partial feature loss when the target is small in UAV image recognition, CutMix is used to perform multi-angle image fusion on the line images captured by UAV, which is used to improve the accuracy of target detection. The improved YOLOX-pruning algorithm model is used for deep learning to prune and sparse the network structure, thus removing the redundant nodes of the network to improve the speed of target detection. In this experiment, manually labeled line images are fed into the model to train the features of the faulty components in the images. With a 50% reduction in channel parameter size and multi-angle feature fusion, the algorithm target detection speed is improved by 2.569 frames per second and the mAP value of the faulty data set is improved by 3.378%, reducing the amount of operation while improving the target detection accuracy.

The smart grid is a system involving the synchronous transmission of information and energy, which is a way to achieve power automation. The development of efficient and clean electric vehicles (EV) is an effective way to reduce oil dependence and alleviate air pollution while simultaneously transforming energy structure and energy consumption mode and improving energy comprehensive utilization efficiency. At present, the load forecasting methods of electric vehicles are mainly divided into two categories. One is to use mathematical models to simulate the charging behavior of electric vehicles so as to obtain the predicted value of electric vehicle load. The other is to use the model in statistical learning to predict based on historical data. The traditional prediction methods of EV charging load include regression analysis, similar day method, etc. Modern prediction methods include prediction methods based on wavelet analysis, prediction methods based on neural network, and support vector machine (SVM). This chapter uses the deep learning model based on historical data to predict the spatio-temporal dynamic load of electric vehicles.

With the increasing number of electric vehicles (EVs) in recent years, road congestion is becoming a common phenomenon, which not only prolongs travel time but also causes anxiety for EV users. Therefore, this paper proposes a time-varying shortest path search method for traffic flow (TF) and establishes an EV route planning model based on this method to plan the optimal path for users. Firstly, a convolutional neural network (CNN) is used to predict TF and a new queuing model is established to calculate the charging queuing time. Then, a path planning model considering mid-way charging is established based on the predicted TF data and charging queuing model. Finally, the performance of the proposed method is tested using road network maps of different scales, and a case study on the optimal path of EV with mid-way charging under the minimum objective function is conducted based on a real traffic network. The results show that the proposed time-varying shortest path search method under the TF network can quickly calculate the optimal path and has great potential for solving practical problems.

In order to defraud state subsidies, some unscrupulous users use improper means to steal photovoltaic (PV) power. This behavior brings potential safety hazards to photovoltaic grid-connected operations. In this paper, a photovoltaic power-stealing identification method based on similar-day clustering and interval prediction of the quantile regression model for long short-term memory neural network (QRLSTM) is proposed. First, photovoltaic data are clustered into three similar days by the similar-day clustering according to weather conditions. Second, compared with the quantile regression neural network (QRNN) prediction method, the good prediction performance of the QRLSTM method is illustrated. Third, using the prediction intervals with different confidence levels on three similar days, according to the time scale (short-term, medium-term and long-term) combined with different electricity-stealing judgment indicators, a three-layer photovoltaic power-stealing screening framework is constructed, and the degree of user power stealing is qualitatively analyzed. Last, the power generation data of eight photovoltaic users in a certain region of northwest China and the data of four groups of artificially constructed power-stealing users are used as an example for simulation. The simulation results prove the feasibility of the proposed method in this paper.

为了提高纯电动客车的安全性,研究了对中央驱动式纯电动客车,在长下坡工况制动盘温升过高的问题.采用模拟仿真和实际道路试验,对某电动客车的制动器和制动卡钳温升进行相关性研究,基于Ansys Workbench软件建立制动盘的热力耦合模型,利用Matlab根据实验数据拟合出制动钳与制动盘温升的相关函数.结果表明:该中央驱动式纯电动客车利用电机反拖持续制动坡度约为3.99%,坡度更大时需要行车制动参与.制动卡钳和制动盘温升相关系数高达0.9,从而,制动卡钳温度可用于监测制动盘热衰退和安全预警.这些结果,可为纯电动客车长下坡制动安全提供数据支撑.

Machine learning (ML)-based methods have attracted great attention in the multi-objective optimization prob-lems, which is the key challenge in the energy management strategy (EMS) of the multi-power hybrid system. Our recently published research in this journal verified the effectiveness and feasibility of a Deep Deterministic Policy Gradient (DDPG)-based EMS in the charge-sustaining (CS) stage of a multi-mode plug-in hybrid vehicle (PHEV). However, the application of ML-based-EMS in the charge-depletion (CD) stage and the regenerative braking mode of PHEV are still missing. This study proposes a discrete-continuous hybrid actions-based hier-archical EMS to optimally distribute the dual-motor driving force in battery electric driving and regenerative braking. In the upper layer of EMS, DDPG is trained to learn the torque distribution principles of dual-motor operation to achieve better energy efficiency without losing dynamic performance. Meanwhile, the total recoverable braking torque is also determined by the upper layer EMS considering the braking demand, me-chanical and electrical braking system conditions, vehicle safety, and the provisions of law. In the lower level of EMS, the driving mode is determined under the guidance of energy consumption optimization. The verified results show that the proposed EMS outperforms other deep reinforcement learning (DRL)-based hierarchical and non-hierarchical EMSs.

The application of electric vehicles (EVs) in the logistics industry has become more extensive. However, the mileage limitation of electric logistics vehicles (ELVs) and the long-distance distribution of ELVs have become urgent problems. Therefore, this paper proposes a long-distance distribution model for ELVs based on dynamic traffic information considering fleet mileage, distribution time and total distribution cost as the optimisation objectives, thus reasonably planning road selection and charging, and alleviating “mileage anxiety” in the long-distance distribution of ELVs. The model proposed in this paper comprehensively considers the characteristics of the high-speed and low-speed roads, the changes in road traffic flow on weekdays and non-weekdays, the time-of-use electricity price of electric vehicle charging stations (EVCSs) and uses the M/M/s queuing theory model to determine the charging waiting time. Finally, a real traffic network is taken as an example to verify the practicability and effectiveness of this model.

In order to help improve the economy, energy savings and emission reductions of pure electric buses, based on the driving data, a new driving cycle construction method is proposed. Through the dividing of short trips and the calculation of characteristic parameter values, two typical driving conditions (weekday driving condition and weekend driving condition) are constructed via principal components analysis and the k-means clustering method, and both have a high degree of compatibility with the actual conditions. Based on the two typical driving conditions, the CRITIC (Criteria Importance Through Intercriteria Correlation) method and the quantitative analysis are used to establish a quantitative evaluation model to score the economy of the driver's driving behavior. The result shows that the weekend working condition with the better traffic environment promotes the generation of aggressive driving behavior and increases the random fluctuation seen in the driver's driving process: for the weekend driving condition, the proportion of low economic efficiency is about 4.5 times bigger than the proportion on weekdays, and the former's fluctuation range for the driving behavior score is 37% higher than that of the latter, meaning that the overall economy of the pure electric bus is much worse on weekends.

为探究即时配送服务(IDS)骑手普遍的危险骑行意图驱动来源,基于计划行为理论(TPB)提出态度、主观规范和直觉行为控制与行为意图的假设关系,拓展安全氛围、薪资期望和工作场所冲突3个变量,构建IDS骑手危险骑行的结构方程模型.研究结果表明:拓展TPB比TPB具有更好的拟合适配度,知觉行为控制、安全氛围和工作场所冲突对配送危险骑行意图的作用更突出;配送骑手普遍缺乏有效的交通安全考核和安全教育培训;加强IDS骑手雇佣关系的解释能力和提高骑手话语权是解决IDS交通安全问题的关键.