Shunt active power filters(SAPF) are employed to improve power quality by injecting compensating harmonic current. The modular SAPF offers many new capabilities that are otherwise unavailable by the conventional SAPF, such as high crest factor compensating current and fast dynamic response. However, there are still challenges that are needed to be addressed, such as resonance and stability issues associated with the modular SAPF. To investigate these issues, this work first presents a mathematical model for conventional modular SAPF system. Based on the mathematical analysis, a new hybrid three-level modular SAPF is presented that is composed of two types of modules, each with different current carrying capacities, LCL filter parameters, and switching frequencies. The proposed hybrid system provides a wider current tracking bandwidth and fast dynamic response as compared to the present modular SAPF. A novel self-adaptive, active damping strategy is proposed that effectively suppresses resonance and coupling between modules. Mathematical analysis and experimental results have been used to verify the proposed system.

Targeting the problem whereby electric vehicle charging loads have large temporal randomness, which affects the accuracy of load prediction, an electric vehicle charging load prediction method based on an improved long short-term memory (LSTM) neural network is investigated. The similarity of EV charging load curves is calculated and the data related to EV charging loads are clustered according to the similarity using a spectral clustering algorithm. The principal component analysis method is used to extract the principal components from the clustering results of EV load data. The LSTM neural network takes the main components of EV charging load as inputs, updates the state of the storage unit through the activation function, introduces an attention mechanism to improve the structure of the network, and outputs the prediction results of the EV charging load through the operation of the input gate, forgetting gate, and output gate. The experimental results show that this method can accurately predict the hourly and daily charging loads of electric vehicles and provide support for their orderly charging of electric vehicles.

As a retailer between the energy suppliers and end users, the integrated energy service provider (IESP) can effectively coordinate the energy supply end and the energy use end by setting energy prices and energy management. Because most of the current research focuses on the pricing of electricity retailers, there are few studies on IESP energy pricing and management, which are still at the initial stage. At the same time, the existing research often does not consider the impact of demand response (DR) and uncertainties, such as natural gas and electricity wholesale prices, on the pricing of IESP. It is necessary to model the DR and uncertainties in the integrated energy system. Aiming at the inadequacy of the existing research and to address the energy pricing and management of IESP, this paper develops a two-stage stochastic hierarchical framework, which comprehensively considers the DR strategy of the user end, characteristics of the electricity/gas/heat storage and the uncertainties of electricity and gas wholesale prices. The proposed hierarchical model for energy pricing and management is a two-layer model: the upper layer is the problem of maximizing the benefits of IESP, and the lower layer is the problem of minimizing the energy cost of user agents. Through the complementary transformation, the linearization method and the strong duality principle in the optimization theory, the model is transformed into a mixed-integer linear programing (MILP) problem, which can be easily solved by the off-shelf commercial solver. Finally, the simulation results are provided to demonstrate the interactive operation between the IESP and user agent through energy prices setting, DR strategy and energy management.

The transient stability assessment based on machine learning faces challenges such as sample data imbalance and poor generalization. To address these problems, this paper proposes an intelligent enhancement method for real-time adaptive assessment of transient stability. In the offline phase, a convolutional neural network (CNN) is used as the base classifier. A model training method based on contrastive learning is introduced, aiming to increase the spatial distance between positive and negative samples in the mapping space. This approach effectively improves the accuracy of the model in recognizing unbalanced samples. In the online phase, when real data with different distribution characteristics from the offline data are encountered, an active transfer strategy is employed to update the model. New system samples are obtained through instance transfer from the original system, and an active sampling strategy considering uncertainty is designed to continuously select high-value samples from the new system for labeling. The model parameters are then updated by fine-tuning. This approach drastically reduces the cost of updating while improving the model’s adaptability. Experiments on the IEEE39-node system verify the effectiveness of the proposed method.

Dual three-phase (DTP) permanent magnet synchronous motors (PMSMs) are specialized machines which are commonly used for high power density applications. These machines offer the merits of high efficiency, high torque density, and superior supervisor fault tolerant capability compared to conventional three-phase AC-machines. However, the electrical structure of such machines is very complicated, and as such, control becomes challenging. In conventional vector controlled DTP-PMSMs drives, the components of the dq-subspace are associated with electromechanical energy conversion, and two currents, i.e., I_d and I_q belonging to this subspace, are used in feedback-loops for control. Such orthodox control methods can cause some anomalies e.g., the voltage source inverter’s (VSI) dead time effect and other nonlinear factors, and can induce large harmonics. These glitches can be greatly alleviated by the introduction of the two-extra current loops to directly control the currents in Z₁Z₂-subspace in order to suppress the insertion of harmonics. In this paper, two approaches—one with two-current loops and other with four-current loops—for vector controlled DTP-PMSMs are investigated with the aid of different MATLAB-based simulations. Furthermore, in the paper, the influence of additional current loops is quantified using simulation-based results.