Against the backdrop of China’s carbon peak and carbon neutrality goals, the energy system is undergoing rapid transformation. As a vital component, virtual power plants (VPPs) have become a key pathway for integrating distributed energy resources (DERs) into power systems and electricity markets. However, VPP operations still face challenges, such as fluctuating renewable energy output, uncertainty in demand response capabilities, and strategic behavior. To address this, a bi-level optimization scheduling method for VPPs is proposed via Stackelberg game and Q-learning-based differential evolution (QDE). First, a two-stage Stackelberg dynamic game model is established for VPP operators and the user side. VPP operators optimize electricity prices and controllable unit outputs, while industrial users and electric vehicles adjust their loads according to electricity prices to maximize flexibility. Secondly, a solution method combining Q-learning-based differential evolution with quadratic programming is proposed. The upper layer employs QDE optimization, while the lower layer is solved via quadratic programming. Q-learning adaptively adjusts crossover and mutation rates to enhance search capability; parameters are determined through orthogonal experiments, with multi-factor variance analysis validating its effectiveness. Thirdly, sensitivity analysis reveals the impact of carbon trading interval lengths on emissions, identifies the optimal benchmark electricity price, and demonstrates the model’s effectiveness under renewable generation uncertainty. Finally, simulation results demonstrate that this approach reduces carbon emissions and EV charging costs by 20.31% and 8.67%, respectively.

We present a new methodology for handling errors of Artificial Intelligence (AI) by introducing weakly supervised AI error correctors with a priori performance guarantees. These AI correctors are auxiliary maps whose role is to moderate the decisions of some previously constructed underlying classifier by either approving or rejecting its decisions. The rejection of a decision can be used as a signal to suggest abstaining from making a decision. A key technical focus of the work is in providing performance guarantees for these new AI correctors through bounds on the probabilities of incorrect decisions. These bounds are distribution agnostic and do not rely on assumptions on the data dimension. Our empirical example illustrates how the framework can be applied to improve the performance of an image classifier in a challenging real-world task where training data are scarce.

Insulator faults are an important factor in causing outages and accidents in power transmission lines. In response to problems related to inefficient insulator positioning, limited robustness of insulator defect feature extraction methods, and the scarcity of defective insulator samples leading to poor classifier generalization, a method for insulator defect detection and recognition based on vision big‐model transfer learning and a stochastic configuration network (SCN) is proposed. First, data augmentation methods, such as Mosaic and Mixup, are employed to mitigate overfitting in the YOLOv7 network. Second, StyleGanv3 adversarial generative networks are used to augment the dataset of defective insulators, which enhances dataset diversity. Third, a vision big‐model transfer learning method based on DINOv2 is introduced to extract features from insulator images. Finally, an SCN classifier is used to determine the status of insulators. Experimental results demonstrate that the applied data augmentation methods effectively mitigate overfitting. YOLOv7 accurately detects insulator positions, and the use of the DINOv2 feature extraction method increases the accuracy of insulator defect recognition by 28.6%. Compared with machine learning classification methods, the SCN classifier achieves the highest accuracy improvement of 17.4%. The proposed method effectively detects insulator positions and recognizes insulator defects.

AI errors pose a significant challenge, hindering real-world applications. This work introduces a novel approach to cope with AI errors using weakly supervised error correctors that guarantee a specific level of error reduction. Our correctors have low computational cost and can be used to decide whether to abstain from making an unsafe classification. We provide new upper and lower bounds on the probability of errors in the corrected system. In contrast to existing works, these bounds are distribution agnostic, non-asymptotic, and can be efficiently computed just using the corrector training data. They also can be used in settings with concept drifts when the observed frequencies of separate classes vary. The correctors can easily be updated, removed, or replaced in response to changes in distributions within each class without retraining the underlying classifier. The application of the approach is illustrated with two relevant challenging tasks: (i) an image classification problem with scarce training data, and (ii) moderating responses of large language models without retraining or otherwise fine-tuning.

Multiphase motors naturally have a higher fault-tolerant capability and therefore their fault tolerant control(FTC) has attracted a wide range of attention. Unfortunately, existing research on phase current derating fault tolerant control (PCDFTC) is only aimed at motors with multiple sets of identical symmetrical stator windings and can only be derated using the entire set of windings as the smallest unit. The torque operation range(TOR) is therefore limited. To overcome the limitation, this paper proposes an online PCDFTC method employing asymmetrical current RMS constraints that can be implemented online to generate current references and achieve the minimum stator copper losses(SCL) in the full TOR while ensuring a circular fundamental rotating magnetic field. This approach solves the problem that the current references based on the offline optimization method are too large to store in the microcontroller. The effectiveness of the proposed method is verified by experimental results.

Dual active bridge is a promising isolated bidirectional DC/DC converter which has attracted more and more attention due to its advantages of zero-voltage switching, great voltage regulation ability, high power density, and fast dynamic response. However, the DC bias problem is seldom mentioned, which may lead to system instability and decreased efficiency. Besides, whether the converter can realize soft switching in the dynamic process is unclear. This paper proposed a novel phase-shift loading mode which can not only realize the DC bias elimination but also achieve zero voltage switching (ZVS) during a transient state. Finally, a 4.5 kW prototype is built and experimental results are provided to verify the conclusion.

The multi-material BOOST inductor which combines multiple powder cores having different material properties can achieve high power density and high efficiency. However, powder cores have a unique feature that the relative permeability varies depending on the magnetic field intensity. The modeling of multi-material powder core inductor is not well discussed in the relevant literature. This paper proposes a novel model of DC superimposition inductance, a novel model of DC superimposition core loss and a model of winding loss respectively. These models apply to all different kinds of powder core materials, core sizes and wire sizes. According to the above models, the BOOST inductor is optimized based on the given material and size database. The optimized inductor has higher power density and higher efficiency compared with the original design.

The explosion in the number of electric vehicles (EVs) has had a significant impact on the energy systems and structures of cities. Large-scale EVs inevitably increase the load on the grid, while uncoordinated vehicle to Grid (V2G) technologies pose challenges to the stability and security of the grid. This paper introduces a global intelligent method to find optimal cooperation charging/discharging strategies for EVs to minimize the operation cost. EVs aggregates co-ordinate the energy information and needs of all EVs and use real-time pricing based on micro-grid loads to influence EV charge-discharge behavior. Particle swarm optimization (PSO) is introduced to solve the EV scheduling problem. This study also discusses the negative impact on the energy system of different strategies for charging EVs. Simulation shows that this smart charging strategy and improved PSO can effectively decrease the operation cost of EVs and reduce the load for each micro-grid.

Due to their inherent fault tolerance capability, multiphase drives are favored in applications with high-reliability requirements. In order to further enhance the fault-tolerant performance, many study efforts have been devoted to different postfault control strategies. In most of the strategies, knowledge of the fault type and location is required prior to the fault-tolerant operation. Therefore, fault detection is an indispensable procedure. Making use of the zero-sequence current constraints, this paper proposes a generalized method to detect and locate the open-phase faults in symmetrical multiphase machines. This method is simple and can be applied to symmetrical multiphase machines with any phase numbers. Experimental results prove the fast detection speed and robustness of the proposed method even in machine transients (such as speed and torque steps) and multiple-open-phase faults.

The energy management of the microgrid (MG) with electric vehicles (EVs) is a large-scale optimization problem where the goal should take into account the performance and economic benefits of the power system while meeting the travel needs of EVs. Due to the development of vehicle to grid (V2G) technologies and demand response (DR), the relationship between EVs and MG becomes currently closer, which leads to a more complex situation. Therefore, the relationship of interest between MG and EVs has to be clarified to improve the performance of MG and EVs to achieve a win-win situation. This paper proposes a two-tier energy management strategy that considers the benefits for both MG and EVs. The first layer ensures the performance of the MG, while the second layer reduces the charging cost from the perspective of the car owners. In addition, based on the existence of uncertain parameters, mixed type variables and nonlinear constraints in the optimization problem, the differential evolution, stochastic search and greedy algorithm are used to analyze and find the optimal solution. Simulation results verify the effectiveness of the proposed strategy and solutions, which benefit both the MG and EV owners.

Dual-active-bridge (DAB) DC-DC isolated converter is widely used in many applications e.g. power supplies, transportations and renewable power systems. In this paper, general modulation characterization and optimization of DAB converters have been analyzed and advanced modulation methods are proposed. Simulations and experiments are implemented to verify the effectiveness of the proposed modulation strategies.

Modular multilevel converter (MMC) is an emerging and highly attractive multilevel topology for medium- and high-voltage applications. This paper proposes an MMC-based power electronic transformer (PET) for dc distribution grid. The high-voltage side is an H-bridge MMC to generate a high-frequency adjustable-magnitude sinusoidal voltage, and the low-voltage side is paralleled multiwindings and rectifiers to allow high-load currents. As the number of MMC voltage levels increases, the complexity of the control system as well as the control algorithm increases largely. In order to simplify the control system and reduce the computation time, a hierarchical control system is designed for the MMC-based PET, and the relevant hierarchical control algorithm is also presented in this paper. The modulation method, capacitor voltage balancing method, and communication solution under the hierarchical control system are all discussed in detail. Simulation and experimental results are presented to demonstrate this system and control method.

Five-level hybrid-clamped (5L-HC) inverter is a newly proposed topology which is suitable for high-performance medium-voltage drives. This paper presents a capacitor voltage balancing strategy for a 5L-HC inverter, including the voltage balancing of DC-link capacitors and flying capacitors. Classic phase-shifted PWM is used to control this converter and a mean value model of flying capacitor and neutral point currents is established. The voltages across the central DC-link and flying capacitors are regulated by adjusting the width of four PWM signals, which varies the operation time of redundant switching states in each switching period essentially. The relationship between the neutral-point currents and the output voltage is also studied and the upper and lower DC-link capacitor voltages are balanced by zero-sequence voltage injection. Experimental results are presented to verify the validity of this strategy.