Among the existing modulation strategies, SVPWM has attracted much attention due to its high DC voltage utilization and good dynamic performance. However, the traditional SVPWM algorithm is computationally complex, has low operational efficiency, which limits its promotion in practical applications. To address these issues, this paper takes the three-phase two-level voltage source converter as the research object and proposes an optimized SVPWM modulation strategy based on simplified basic vectors. The effectiveness of this strategy in optimizing SVPWM and maintaining a single power factor operation on the AC side is verified through simulation analysis.

Traditional multi-port converters have problems such as low integration, low efficiency and complex control, while non-isolated converters have become the focus of research due to their simple structure, high efficiency and easy scalability. In this paper, a non-isolated three-port converter with variable port structure is proposed, which realizes flexible switching between Buck, Boost and Buck/Boost modes by controlling the switch, weakens the voltage constraint and expands the operating range. Based on the time scale, the control strategy is divided into three layers: modulation optimization, mode switching and variable structure switching, and the corresponding strategies are designed to optimize the performance, achieve smooth switching, and further expand its operation range.

This paper proposes a flexible distribution network operation optimization strategy considering mobile energy storage system (MESS) integration. With the increasing penetration of renewable energy in power systems, its stochastic and intermittent characteristics pose significant challenges to grid stability. This study introduces an MESS, which has both spatial and temporal controllability, and soft open point (SOP) technology to build a co-scheduling framework. The aim is to achieve rational power distribution across spatial and temporal scales. In this paper, a case study uses a regional road network in Chengdu coupled with an IEEE 33-node standard grid, and the model is solved using the non-dominated sorting genetic algorithm III (NSGA-III) algorithm. The simulation results show that the use of the MESS and SOP co-dispatch in the grid not only reduces the net loss and total voltage deviation but also obtains considerable economic benefits. In particular, the net load peak-to-valley difference is reduced by 20.1% and the total voltage deviation is reduced by 52.9%. This demonstrates the effectiveness of the proposed model in improving the stability and economy of the grid.

This paper proposes a multivariate cooperative power flow control strategy based on multi-port converter to address the power fluctuation problem of grid-connected renewable energy. By integrating the power interaction of grid sub-grids and port-level model predictive control through a hierarchical control architecture, we realize the cooperative optimal scheduling of distributed power sources, energy storage and loads, and design an adaptive smoothing mechanism that combines real-time data and dynamic optimization algorithms to regulate the power flow, effectively suppressing the fluctuations of renewable energy sources and the sudden changes of loads.

Deep learning-based algorithms are considered an efficient solution to carry out the insulator defect diagnosis task based on the aerial images captured by Unmanned Aerial Vehicles (UAVs) for electric power systems. However, the sufficient and accurate annotations of image samples required by deep learning-based models can be costly or not feasible in practice. This paper proposed an active learning-based solution for insulator defect diagnosis of electric transmission networks. The proposed solution aims to identify the most valuable samples and assign ground-true labels to significantly reduce the labeling effort. Specifically, Batch Active learning by Diverse Gradient Embedding (BADGE) strategy is adopted for sampling and GradCAM++ is used to extract the key regions of image samples iteratively. Then, a region-sample pair construction method is proposed during the labeling stage enabling the model to focus on the most discriminative regions based on a well-designed loss function. The proposed solution is extensively assessed through experiments and the results demonstrate that F1 scores of four popular CNN models trained with 1/3 of the total samples can be increased by up to 2.0% compared to the fully-labeled baseline solutions.

Loop closing operation of distribution network is necessary to realize load transfer without power supply, but it may generate large loop closing current and lead to load transfer failure. In order to minimize the impact of loop closing current on load transfer, a control strategy is proposed to solve the problem of voltage overrun by adjusting the loop closing current, taking into account the change of the electrical quantity in the distribution network after accessing the distributed generator (DG), which will lead to the success of loop closing. By analyzing the generation mechanism and influencing factors of the loop closing current in the loop closing process, the electrical quantities affecting the loop closing current in the loop closing process are quantified, and then the regulation method of the loop closing current is constructed, and the regulation strategy of the loop closing current in distribution networks based on particle swarm optimization (PSO) is proposed. Examples show that the method can maximize the win-win situation of loop closing operation and distributed power supply, and effectively improve the security and reliability of the loop closing process.

Detecting distribution network insulators with deep learning algorithms is challenging due to the complex background and the small size of the targets. Since Visual Foundation Models(VFM) have shown impressive performance in general domains, this paper attempts to utilize the features extracted from VFM to assist YOLOv8 in achieving more accurate detection. InternImage is selected to provide the general visual priors, and its features are combined with the feature maps of YOLOv8 after passing through a feature fusion block. The proposed model is evaluated through extensive experiments by training on the distribution network insulator dataset and demonstrated to be superior to the original YOLOv8 in this task. Further, the superiority of the proposed model in detecting small-size insulators is illustrated by visualization. Additionally, the effectiveness of the proposed solution in identifying small-size insulators is demonstrated through visualization.

Distributed photovoltaic (PV) access to the regional power grid is the future trend, to enhance the regional power grid distributed PV carrying capacity is a key technical issue in the future planning and operation of the regional power grid. Aiming at the problem that it is difficult to directly calculate the distributed photovoltaic carrying capacity of large-scale regional power grid, a distributed photovoltaic carrying capacity calculation method is proposed under the grid-based management of regional power grid. First, the extreme scenarios and capacity caps for distributed PV outlets in the regional power grid are determined by considering the timing characteristics of distributed PV and roof distribution characteristics. Secondly, based on the mixed integer second-order cone programming, a variety of active management measures are added to establish the calculation model of the distributed photovoltaic carrying capacity of the regional power grid unit. Finally, based on the dynamic reconfiguration principle of the grid regional power grid, a calculation model of the distributed photovoltaic carrying capacity of the regional power grid is established. Case analysis shows that the proposed method improves the carrying capacity of distributed PV in the regional power grid by 11.82%, which proves the effectiveness and feasibility of the proposed method

As energy consumption and environmental problems become more and more prominent, IESs are developing rapidly as a way to efficiently use and optimize energy. Meanwhile, hydrogen energy can provide important support and development space for the efficient and stable operation of IESs. In order to improve resource utilization, this paper proposes a day-ahead scheduling model of IES considering hydrogen storage to participate in frequency regulation auxiliary services. The optimal scheduling model proposed in this paper takes the maximum day-ahead revenue as the target. It considers the constraints of hydrogen storage tanks, fuel cells, and other members to realize the day-ahead scheduling of the IES participating in the frequency regulation market. In the end, the effectiveness of the established model is verified by arithmetic cases.

With the continuous development of State Grid middle platforms and expansion of power grid applications, SG-CIM model verification has received increasing attention. It plays a crucial role in ensuring the data accuracy and consistency for information applications tailored to the needs of different professional departments in the power industry. This paper provides an overview of the SG-CIM middle platform, emphasizing its crucial role in supporting data sharing and analysis applications between different professional and operational units. We outline key steps involved in the verification and validation process for SG-CIM models, including input checks, model validation, sensitivity analysis, and model evaluation modules. Furthermore, the paper also highlights how this framework fosters collaboration between different parts of an organization by ensuring that data used in these applications is accurate and consistent across different business processes.

This paper is about a real-time parallel computing technology in LAN. The main contents are as follows: the modular management of power system based on power grid topology, and the corresponding calculation mode of transmission and distribution between systems are defined; And the directed graph based on the correlation between various transmission and distribution calculation methods; The nodes in the target directed graph correspond to the calculation equation of the State Grid, and all the edges in the overall target directed graph correspond to the correlation relationship between the calculation equation of the State Grid; Then, according to the parallel calculation model established in the general objective directed graph, the parallel calculation module is used to realize the parallel calculation of the calculation formula of the national grid, and the corresponding results are obtained. Through this technology, the computing quality can be greatly improved, so as to meet the real-time requirements of power system dispatching.

With the continuous accumulation of large-scale power grid data, the traditional centralized data analysis method is more and more expensive for data transmission. Based on this, we designed a grid big data monitoring and analysis system and transferred the computation process to the edge node close to the data source through an edge computing strategy. On the one hand, data processing and data analysis algorithms are encapsulated by container technology, and the algorithm is mirrored to the edge nodes of the power network through the system to complete the computation. On the other hand, the computing clusters are deployed at the edge nodes of the power network, which is responsible for the scheduling, execution, and status monitoring of computing tasks. Computing tasks can be flexibly managed in a cluster by extending user-defined resources. Through the reserved parameters, users can intervene in task execution policies, and tasks can be configured. The edge node sends the calculation result or early warning information to the central monitoring service through the asynchronous message. Compared with the traditional centralized data analysis system, the proposed method relieves the problem of the overhead of massive data transmission in the network, reduces the application cost, helps to apply the data analysis to more edge side nodes, and fully excavates the potential value of grid data.

In order to solve the problems of order reduction and customer churn caused by sorting delays, the author proposes a method for e-commerce sorting equipment based on cloud computing. This method mainly adopts double-layer sorting equipment; compared with single-layer automatic sorting equipment, double-layer sorting equipment has the characteristics of higher efficiency and smaller floor space. The sorting method adopts the "group sorting" method, which can effectively improve the sorting efficiency of the sorting equipment. The algorithm method adopts the mathematical model based on cloud computing for calculation. Experimental Results. The author adopts the cloud computing-based "composition sorting" double-layer sorting equipment; compared with the traditional single-layer sorting equipment, the throughput of the single-layer sorting strategy is 0.72 pieces/s when the conveying speed of the conveyor belt is the same, the throughput of the two-layer same-direction strategy is 1.46 pieces/s, the throughput of the balanced load strategy is 1.97 pieces/s, and the throughput of the group sorting load strategy is 2.57 pieces/s. This method can effectively solve the problems of order reduction and customer churn caused by sorting delays.

Flexible on-grid/off-grid microgrids can improve the resilience of distribution grids and achieve autonomous management of distributed energy. The economic dispatch of microgrid considering the uncertainty of renewable energy can fully absorb distributed energy, ensure grid security, improve power quality, and reduce operating costs. In order to realize multi-time-scale robust economic dispatch of microgrid considering discrete events, the microgrid dispatching models and results using different uncertainty analysis methods are compared and analyzed. It is proved that the data-driven distributed robust optimization method can take into account the model’s economy and conservatism at the same time. Aiming at the improvement of Day-ahead-Intraday-Real-time forecast accuracy of renewable energy and load, based on the model predictive control theory, a rolling scheduling strategy for multi-time scale collaborative optimization of microgrid is designed. The proposed optimal scheduling strategy can effectively deal with discrete events in the microgrid and realize the safe and economical operation of the microgrid.

Compared with independent scheduling, electro-thermal coupling scheduling can effectively reduce operating costs and improve renewable energy consumption. Based on the unified energy path theory, the heating network is analyzed and modeled. Based on the conservation of momentum, mass and energy, the water path model and heat path model of the heating network are constructed. Derive the water equations of the heating network as well as the equations of the heat circuits. Establish an optimal scheduling model for electrothermal coupling multi-energy flow system, using the unified energy flow method to calculate the heating network can be unified with the power system in multiple time scales. The results show that the optimal scheduling model established in this paper can improve the scheduling flexibility and solve accuracy.

Substation centralized medium voltage distribution network (SCDN), makes the large-scale complex medium voltage network into a dynamic radial network with a fixed static topology and a central substation as a unique power supply source. A novel concept of "Supply Region of substation centralized medium voltage distribution network (S-SCDN)" is proposed to generate a "one map" as the core technology of the "urban distribution network brain". The prototype of SCDN and S-SCDN and single line diagram prototype are proposed, and their basic characteristics are studied. Then, the application scenario analysis based on S-SCDN single line diagram including a situation awareness dynamic visualization platform under SCDN/S-SCDN concepts is also proposed, which indicates that S-SCDN single line diagram provides a new analysis tool for urban complex distribution network with a wide application.

This paper uses big data technology to mine distribution network operation rules from massive power big data, and establishes an accurate early warning model for distribution network changes to predict the occurrence of changes such as overload, low voltage and other abnormal events in advance. In this paper, a switch abnormality sensing and adaptive correction algorithm is proposed to realize the sensing of tie switch abnormality loop closure, locate the optimal loop solution point and program the loop solution. The study found that the distribution network model management of the change order can partially solve the problems of the accuracy and timeliness of the distribution network information integration. This system can meet the requirements of atomicity and integrity of model management.