Power transmission and distribution system electric energy quality real-time monitoring and dynamic regulation and control system and method

Abstract

The invention discloses a power transmission and distribution system electric energy quality real-time monitoring and dynamic regulation and control system and method, the system comprises a multi-dimensional heterogeneous data acquisition and access unit, an edge side data processing unit and a feature extraction unit, data preliminary processing and feature extraction are locally completed through an edge calculation model, and transmission delay is reduced; and applying the optimized attention mechanism model to deeply mine correlation characteristics among power quality parameters such as voltage amplitude, frequency and the like, and accurately evaluating and predicting the power quality state. And according to the evaluation prediction result and the system operation constraint, a regulation strategy is dynamically generated, and after equipment execution, the result is fed back for optimizing subsequent evaluation. According to the invention, a whole-process closed loop from data acquisition, analysis to regulation is realized, data processing efficiency and electric energy quality analysis regulation precision are effectively improved, and stable operation of a power transmission and distribution system and reliable electric energy quality are guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of power quality monitoring of power transmission and distribution systems, and in particular to a system and method for real-time monitoring and dynamic regulation of power quality of power transmission and distribution systems. Background Art

[0002] With the advancement of smart grid construction, the demand for real-time power quality monitoring and dynamic control in transmission and distribution systems has surged. Under traditional centralized data processing models, massive amounts of power quality data are transmitted back to the central cloud, which not only causes network congestion and increased latency, but also creates enormous data processing pressure and security risks. Existing monitoring and control technologies often rely on single models, making it difficult to accurately capture the complex relationships between multidimensional power quality parameters such as voltage amplitude, frequency, and harmonic content. Control strategies also lack dynamic adaptability and are unable to meet the changing operating conditions of transmission and distribution systems.

[0003] Current power quality monitoring and control technologies for transmission and distribution systems face a dual dilemma. First, data processing efficiency is low. Due to the numerous data collection points in the system, the transmission of heterogeneous data to the central server results in lengthy processing, making it impossible to quickly respond to power quality issues. Second, analysis and control accuracy is insufficient. Traditional models struggle to deeply explore the potential connections between power quality parameters, resulting in inaccurate predictions of issues such as voltage sags and swells, and three-phase imbalance. Control strategies lack specificity, making it difficult to ensure stable power quality. Summary of the Invention

[0004] In order to overcome the shortcomings and deficiencies of the prior art, the present invention provides a real-time monitoring and dynamic control system for power quality of a power transmission and distribution system.

[0005] The technical solution adopted by the present invention is a real-time monitoring and dynamic control system for power quality of a power transmission and distribution system, comprising: Multi-dimensional heterogeneous data acquisition and access unit: used to collect multi-dimensional heterogeneous data from the power transmission and distribution system, including voltage amplitude, frequency, three-phase imbalance, harmonic content, voltage sag and swell, and power quality parameters, and access the collected data into the system through an adapted communication protocol; Edge-side data processing and feature extraction unit: Receives data transmitted by the multi-dimensional heterogeneous data acquisition and access unit, performs preliminary data processing using the edge computing model, and extracts feature data related to power quality; Optimize the attention mechanism feature analysis unit: obtain the feature data output by the edge-side data processing and feature extraction unit, analyze the feature data through the optimized attention mechanism model, and mine the correlation features between the data; Power quality status assessment and prediction unit: Based on the features analyzed by the optimized attention mechanism feature analysis unit, it assesses the power quality status of the transmission and distribution system and predicts the power quality status using the edge computing model and the optimized attention mechanism model; Dynamic control strategy generation unit: Based on the results of the power quality status assessment and prediction unit and combined with the operating constraints of the transmission and distribution system, it generates a dynamic control strategy for power quality issues; Control instruction execution and feedback unit: receives the control strategy generated by the dynamic control strategy generation unit, sends the control instruction to the control equipment of the transmission and distribution system for execution, and feeds back the execution result to the power quality status assessment and prediction unit.

[0006] Furthermore, the edge-side data processing and feature extraction unit uses an edge computing model to process data. The specific formula is: ,in, Represents the extracted feature data; Represents the original data collected by the multi-dimensional heterogeneous data collection and access unit; For edge computing models; It is a parameter of the edge computing model, which is obtained by training the historical power quality data of the transmission and distribution system. It is used to control the way the edge computing model processes the original data and extract characteristic data related to the voltage amplitude change trend and frequency fluctuation characteristics of the power quality from the original data.

[0007] Furthermore, the optimized attention mechanism model formula adopted by the optimized attention mechanism feature analysis unit is: ,in, Represents feature data after analysis by the optimized attention mechanism; Feature data extracted by edge-side data processing and feature extraction units; The weight matrix of the attention mechanism model, whose dimension is set according to the dimension of the feature data, is used to adjust the weight of different feature data in the analysis process; It is a bias vector used to enhance the expressiveness of the model. The model learns the correlation between voltage, frequency, and harmonic power quality parameters in the transmission and distribution system, and highlights the characteristic data that have an important impact on power quality assessment and prediction.

[0008] Furthermore, the power quality status assessment and prediction unit adopts an edge computing model and an optimized attention mechanism model to predict the power quality status, and the formula is: ,in, Indicates predicted power quality status data; Feature data extracted by edge-side data processing and feature extraction units; An optimized attention mechanism model; Optimize the weight matrix of the attention mechanism model for the prediction stage, and adjust it according to the historical power quality status data and the prediction target; Optimize the bias vector of the attention mechanism model for the prediction stage; For edge computing models; To predict the parameters of the edge computing model in the prediction stage, the formula predicts the power quality status of the transmission and distribution system in the future by learning the historical voltage sag and three-phase imbalance power quality parameters and combining them with the current characteristic data.

[0009] Furthermore, the formula for generating the control strategy by the dynamic control strategy generation unit is: ,in, Represents the generated dynamic control strategy; Power quality status data predicted by the power quality status assessment and prediction unit; Represents the operating constraints of the transmission and distribution system, including equipment capacity limitations and safe operating range; A policy generation model is built based on an edge computing model and an optimized attention mechanism model; The parameters of the strategy generation model are obtained by learning historical control strategies and system responses under different power quality states, and are used to generate control strategies based on the predicted power quality state and system operation constraints.

[0010] Furthermore, after sending the control instruction to the control device, the control instruction execution and feedback unit adopts an edge computing model for feedback processing of the execution result, and the formula is: ,in, Indicates the feedback data of the executed result after processing; Indicates the actual execution information of the control instruction executed by the device, including execution time and execution status; For edge computing models; The parameters of the edge computing model in this processing process are obtained by learning historical execution results and system status, and are used to extract valuable feedback data for power quality status assessment from device execution information.

[0011] Furthermore, when the multi-dimensional heterogeneous data acquisition and access unit collects data, an edge computing model is used for time synchronization processing of different types of data, and the formula is: ,in, Indicates the timestamp after time synchronization; Indicates the original timestamp when the data was collected; For edge computing models; It is a time synchronization parameter determined based on the clock deviation and communication delay factors of the data acquisition equipment in the transmission and distribution system. It is used for accurate time synchronization of multi-dimensional heterogeneous data such as voltage, current, and frequency.

[0012] Furthermore, in the optimized attention mechanism feature analysis unit, the optimized attention mechanism model introduces an adaptive weight adjustment formula when processing feature data: ,in, Indicates the The attention weight of feature data; Indicates the Characteristic data; 、 where [1] is the query vector and key vector obtained by training the power quality feature data of the transmission and distribution system; Score is a function that calculates the similarity between the feature data and the query vector and key vector. This formula analyzes the correlation between the voltage harmonic content and three-phase imbalance feature data, adaptively adjusts the attention weights of different feature data, and enhances the ability to extract key features.

[0013] Furthermore, the power quality status assessment and prediction unit uses the following assessment model formula when performing status assessment: ,in, Indicates the power quality status assessment result; is the number of features involved in the evaluation; For the The weight of each feature in the evaluation is obtained by learning historical power quality data through an optimized attention mechanism model; Indicates the The formula combines multiple power quality characteristic data such as voltage amplitude, frequency deviation, and harmonic distortion rate to quantitatively evaluate the power quality status of the transmission and distribution system.

[0014] This application also discloses a method for real-time monitoring and dynamic control of power quality in a power transmission and distribution system, comprising the following steps: S1. The multi-dimensional heterogeneous data acquisition and access unit collects multi-dimensional heterogeneous data on voltage amplitude, frequency, three-phase imbalance, and harmonic content from various power quality monitoring points distributed at different locations in the power transmission and distribution system according to a predetermined sampling frequency and communication protocol, and transmits the collected data to the edge-side data processing and feature extraction unit. S2, the edge-side data processing and feature extraction unit receives the transmitted data, uses the edge computing model to perform format conversion and outlier detection on the data, extracts feature data closely related to power quality from the processed data based on pre-set feature extraction rules, and passes it to the optimized attention mechanism feature analysis unit; S3. After the optimized attention mechanism feature analysis unit obtains the feature data, it uses the optimized attention mechanism model to conduct an in-depth analysis of the feature data. By calculating the correlation between the feature data, it highlights the feature information that is important for power quality assessment and prediction, and sends the analyzed feature data to the power quality status assessment and prediction unit. S4. The power quality status assessment and prediction unit uses the edge computing model and optimized attention mechanism model based on the received feature data, combined with historical power quality data and system operating parameters, to assess the power quality status of the current transmission and distribution system and predict the power quality status in the future. The assessment and prediction results are transmitted to the dynamic control strategy generation unit. S5. The dynamic control strategy generation unit generates a dynamic control strategy for power quality issues based on the received power quality status assessment and prediction results and the current operating constraints of the transmission and distribution system using a strategy generation algorithm based on an edge computing model and an optimized attention mechanism model, and sends the control strategy to the control instruction execution and feedback unit. S6. The control instruction execution and feedback unit receives the control strategy, converts the control instruction into a control signal suitable for execution by the control equipment of the power transmission and distribution system, and sends it. It monitors the execution status in real time during the process of the equipment executing the control instruction. After the execution is completed, the execution result is processed and fed back to the power quality status assessment and prediction unit for the next round of power quality status assessment and control strategy generation.

[0015] Beneficial effects: The present invention proposes a real-time monitoring and dynamic control system for power quality in a power transmission and distribution system. The system effectively solves the shortcomings of traditional technologies in data processing efficiency and analysis and control accuracy through the deep integration of edge computing and optimized attention mechanism. In terms of data processing, the system sinks the computing power to the edge side close to the data source, and after collecting multi-dimensional power quality data in the power transmission and distribution system in real time, it completes the preliminary processing and feature extraction of the data directly on the edge, avoiding the network congestion and processing delays caused by a large amount of data being transmitted back to the cloud center, significantly improving the timeliness of data processing, and being able to respond quickly to power quality issues. At the level of analysis and control accuracy, the system uses an optimized attention mechanism model to deeply explore the complex correlations between power quality parameters such as voltage amplitude, frequency, and harmonic content, and accurately identify feature information that plays a key role in power quality assessment and prediction. Based on the results of these in-depth analyses, the system can integrate historical data with the current operating status to accurately assess and predict the power quality status. At the same time, combined with the actual operating constraints of the transmission and distribution system, a highly targeted and dynamically adaptive control strategy is generated, and the precise execution and effective feedback of the control instructions are ensured, forming a complete closed loop from monitoring, analysis, prediction to control, realizing high-precision dynamic control of power quality, ensuring the stable and reliable operation of the transmission and distribution system, and providing strong support for the construction of smart grids. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A diagram of a system unit of the present invention; Figure 2 It is a flow chart of the system operation steps of the present invention. DETAILED DESCRIPTION

[0017] It should be noted that, unless there is a conflict, the embodiments in this application and the features described in the embodiments can be combined with each other. The application is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] like Figure 1 As shown, a real-time monitoring and dynamic control system for power quality of a power transmission and distribution system includes: Multi-dimensional heterogeneous data acquisition and access unit: used to collect multi-dimensional heterogeneous data from the power transmission and distribution system, including power quality parameters such as voltage amplitude, frequency, three-phase imbalance, harmonic content, voltage sag and swell, and access the collected data into the system through an adapted communication protocol; Specifically, the multi-dimensional heterogeneous data acquisition and access unit is the source of information for the entire system. Its core responsibility is to comprehensively and accurately collect various types of power quality-related data in the transmission and distribution system, and effectively connect them to the system to provide a basis for subsequent processing and analysis.

[0019] In power transmission and distribution systems, power quality parameters are numerous and complex. The voltage amplitude, which this unit collects, reflects the voltage level in the power system. Its stability is directly related to the normal operation of various electrical equipment. Excessively high or low voltage amplitudes can cause equipment damage or malfunction. Frequency parameters are crucial indicators of power system operation. The standard frequency of my country's power system is 50 Hz, and frequency fluctuations can affect the performance and lifespan of both generating and consuming equipment. Three-phase imbalance reflects the degree of imbalance in voltage or current between phases in a three-phase system. Excessive imbalance increases line losses and reduces equipment utilization. Harmonic content refers to the amount of frequency components other than the fundamental frequency in the power system. Harmonics can cause equipment heating and interfere with communication systems. Voltage sags and swells describe sudden drops or increases in voltage over short periods of time, which can severely impact sensitive equipment. This multi-dimensional, heterogeneous data is collected through a variety of sensors and monitoring devices, and reliable data transmission and access to the system is achieved using compatible communication protocols, such as Modbus and TCP / IP.

[0020] This unit is of great significance. It ensures that the system can obtain real-time, first-hand data on the operation of the power transmission and distribution system, providing an accurate and comprehensive information foundation for subsequent monitoring, evaluation, prediction, and control of power quality. Only when the collected data is accurate and complete can the subsequent work based on this data be carried out effectively, thus ensuring the stable operation of the power transmission and distribution system and the reliability of power quality.

[0021] Edge-side data processing and feature extraction unit: Receives data transmitted by the multi-dimensional heterogeneous data acquisition and access unit, performs preliminary data processing using the edge computing model, and extracts feature data related to power quality; Specifically, the edge-side data processing and feature extraction unit receives data transmitted by the multi-dimensional heterogeneous data collection and access unit, performs preliminary processing and key feature extraction on the data at the edge, and is an important link in transforming data from its original state into usable information.

[0022] After receiving the raw data, the unit uses the edge computing model to first convert the data format, unifying data from different sources and formats into a standard format that the system can recognize and process for subsequent analysis. At the same time, the data is detected for outliers, identifying and eliminating abnormal data caused by sensor failures, communication interference, etc., to ensure the accuracy and validity of the data. On this basis, key feature data is extracted from the processed data according to pre-set feature extraction rules related to power quality. For example, features such as the change trend and fluctuation range are extracted from the voltage amplitude data, and features such as the frequency change rate and frequency deviation are extracted from the frequency data. These features can more intuitively reflect the status and change patterns of power quality, providing more targeted data for subsequent in-depth analysis.

[0023] The presence of edge-side data processing and feature extraction units significantly reduces data processing pressure on the central cloud and improves the real-time nature of data processing. By completing preliminary data processing and feature extraction at the edge, the transmission of large amounts of raw data is reduced, lowering network transmission costs and latency. This enables the system to respond more quickly to power quality issues, improving overall system efficiency and reliability.

[0024] Optimize the attention mechanism feature analysis unit: obtain the feature data output by the edge-side data processing and feature extraction unit, analyze the feature data through the optimized attention mechanism model, and mine the correlation features between the data; Specifically, after the optimized attention mechanism feature analysis unit obtains the feature data output by the edge-side data processing and feature extraction unit, it uses the optimized attention mechanism model to analyze these feature data and mine the hidden correlation features between the data, which is a key step in achieving accurate power quality assessment and prediction.

[0025] The various parameters of power quality in transmission and distribution systems are not independent of each other, but rather complex interrelationships exist. This unit uses an optimized attention mechanism model to conduct in-depth analysis of extracted feature data, calculating the correlation between different feature data and identifying key feature information that has a significant impact on power quality assessment and prediction. For example, when analyzing the relationship between voltage amplitude and harmonic content, the model can highlight the mutual influence between the two under certain operating conditions and identify the potential impact of changes in harmonic content on voltage amplitude stability. In this way, the unit can filter key features from a large amount of feature data, enhance understanding and grasp of the essence of power quality issues, and provide more valuable information for subsequent status assessment and prediction.

[0026] The application of optimized attention mechanism feature analysis units significantly enhances the system's ability to analyze power quality data. Compared to traditional methods, it can more accurately mine complex relationships between data, avoid missing key information, and provide a more comprehensive and in-depth analysis of power quality status. This lays a solid foundation for accurate assessment and prediction of power quality status, thereby improving the system's ability to warn and address power quality issues.

[0027] Power quality status assessment and prediction unit: Based on the features analyzed by the optimized attention mechanism feature analysis unit, it assesses the power quality status of the transmission and distribution system and predicts the power quality status using the edge computing model and the optimized attention mechanism model; Specifically, the power quality status assessment and prediction unit uses the feature data obtained by the optimized attention mechanism feature analysis unit, and comprehensively uses the edge computing model and the optimized attention mechanism model to assess the power quality status of the transmission and distribution system and predict the future power quality status, which is an important basis for the system to achieve active regulation.

[0028] In terms of power quality status assessment, this unit quantitatively evaluates the power quality status of the current transmission and distribution system based on extracted key feature data and pre-set assessment standards and rules. For example, based on characteristic data such as voltage amplitude, frequency deviation, and harmonic distortion rate, it determines whether the current power quality meets national standards and user needs, and classifies the power quality status into different levels such as good, fair, and poor, clearly and intuitively reflecting the current operating status of the system. In the prediction phase, the unit uses edge computing models and optimized attention mechanism models, combined with historical power quality data and system operating parameters, to learn and analyze data to predict the changing trends of the power quality status of the transmission and distribution system over the next period of time. For example, it predicts whether voltage sags will occur in the next few hours, whether the three-phase imbalance will worsen, and so on, to grasp the changing trends of power quality in advance.

[0029] The power quality status assessment and prediction unit provides a crucial basis for dynamic system control. By accurately assessing the current status and predicting future trends, the system can proactively formulate and implement appropriate control strategies to prevent or mitigate power quality issues. This enables proactive management and optimization of the power quality of the transmission and distribution system, improving power supply reliability and quality.

[0030] Dynamic control strategy generation unit: Based on the results of the power quality status assessment and prediction unit and combined with the operating constraints of the transmission and distribution system, it generates a dynamic control strategy for power quality issues; Specifically, the dynamic control strategy generation unit generates a dynamic control strategy for power quality issues based on the results of the power quality status assessment and prediction unit and combined with the operating constraints of the transmission and distribution system. It is the core link for the system to achieve power quality optimization.

[0031] When the power quality status assessment and prediction unit identifies a current power quality issue or predicts a potential future problem, the dynamic control strategy generation unit begins operation. It first analyzes the specific type and severity of the power quality issue, such as low voltage amplitude or excessive harmonic content. Then, based on the current operating constraints of the transmission and distribution system, including equipment capacity limitations, safe operating range, and system load conditions, a strategy generation algorithm constructed using an edge computing model and an optimized attention mechanism model generates a dynamic control strategy. For example, if the system has adjustable reactive power compensation equipment, the strategy generation unit may generate a strategy to adjust the reactive power compensation equipment output to increase the voltage amplitude. If the harmonic content exceeds the standard, a strategy may be generated to activate filtering devices to reduce the harmonic content.

[0032] The control strategies generated by the dynamic control strategy generation unit are highly targeted and adaptable. They can flexibly formulate control plans based on different power quality issues and system operating conditions, effectively resolving power quality issues while ensuring safe and stable system operation, and achieving dynamic optimization and precise control of power quality in the transmission and distribution system.

[0033] Control instruction execution and feedback unit: receives the control strategy generated by the dynamic control strategy generation unit, sends the control instruction to the control equipment of the transmission and distribution system for execution, and feeds back the execution result to the power quality status assessment and prediction unit.

[0034] Specifically, the control instruction execution and feedback unit receives the control strategy generated by the dynamic control strategy generation unit, sends the control instruction to the relevant equipment of the transmission and distribution system for execution, and feeds back the execution results to the power quality status assessment and prediction unit. It is a key link in realizing closed-loop control of the system.

[0035] After receiving the control strategy, the unit first converts the control strategy into a control signal suitable for execution by the relevant equipment of the transmission and distribution system. For example, for the strategy of transformer tap adjustment, it converts it into the corresponding tap adjustment control signal. Then, the control signal is sent to the corresponding equipment through a reliable communication method to ensure that the equipment can accurately receive and execute the control instructions. During the process of the equipment executing the control instructions, the control instruction execution and feedback unit monitors the execution status of the equipment in real time, including execution time, execution progress and other information, to ensure the smooth execution of the instructions. When the equipment completes the execution, the unit collects and processes the execution results and feeds the processed execution results back to the power quality status assessment and prediction unit. For example, whether the feedback voltage amplitude is increased to the expected range, whether the harmonic content is reduced to within the standard value, and other information.

[0036] The control command execution and feedback unit implements closed-loop control of the system, enabling the system to promptly adjust subsequent evaluation and control work based on the effectiveness of the control strategy. Through continuous monitoring, control, and feedback, the system can continuously optimize power quality, forming a self-improving and self-optimizing operating mechanism to ensure that the transmission and distribution system always maintains a high level of power quality.

[0037] The multi-dimensional heterogeneous data acquisition and access unit is connected to the edge side data processing and feature extraction unit, the edge side data processing and feature extraction unit is connected to the optimized attention mechanism feature analysis unit, the optimized attention mechanism feature analysis unit is connected to the power quality status assessment and prediction unit, the power quality status assessment and prediction unit is connected to the dynamic control strategy generation unit, the dynamic control strategy generation unit is connected to the control instruction execution and feedback unit, and the control instruction execution and feedback unit is connected to the power quality status assessment and prediction unit.

[0038] Preferably, the edge-side data processing and feature extraction unit uses an edge computing model to process data, and the specific formula is: ,in, Represents the extracted feature data; Represents the original data collected by the multi-dimensional heterogeneous data collection and access unit; For edge computing models; This parameter is obtained by training the historical power quality data of the transmission and distribution system. It is used to control the way the edge computing model processes the original data and extract characteristic data related to power quality such as voltage amplitude change trend and frequency fluctuation characteristics from the original data.

[0039] Specifically, the edge computing model of the edge-side data processing and feature extraction unit implements parameterized control over the processing method of the original data, thereby realizing the function of extracting power quality characteristics from multi-dimensional heterogeneous data. The parameters of this model are obtained through historical data training, and can adaptively capture key information such as voltage amplitude change trends and frequency fluctuation characteristics. In practical applications, for the common voltage sag problem in the transmission and distribution system, the model can quickly identify potential power quality risks by analyzing the rate of change and fluctuation range of the voltage amplitude, providing a data basis for subsequent status assessment. This edge-side data processing method effectively reduces the amount of data transmission and improves the system response speed.

[0040] Preferably, the optimized attention mechanism model formula adopted by the optimized attention mechanism feature analysis unit is: ,in, Represents feature data after analysis by the optimized attention mechanism; Feature data extracted by edge-side data processing and feature extraction units; The weight matrix of the attention mechanism model, whose dimension is set according to the dimension of the feature data, is used to adjust the weight of different feature data in the analysis process; It is a bias vector used to enhance the expressiveness of the model. The model learns the correlation between power quality parameters such as voltage, frequency, and harmonics in the transmission and distribution system, and highlights the characteristic data that have an important impact on power quality assessment and prediction.

[0041] Specifically, the optimized attention mechanism model achieves differentiated attention to different feature data through the design of weight matrices and bias vectors. When processing multi-dimensional parameters such as voltage, frequency, and harmonics, the model can automatically adjust attention weights by learning the correlations between these parameters, highlighting features that have a significant impact on power quality assessment and prediction. For example, when analyzing harmonic pollution issues, the model can improve the accuracy of harmonic source location and impact assessment by increasing attention to the correlation between harmonic content and voltage distortion rate, providing more precise decision-making basis for harmonic control.

[0042] Preferably, the power quality status assessment and prediction unit adopts an edge computing model and an optimized attention mechanism model to predict the power quality status, and the formula is: ,in, Indicates predicted power quality status data; Feature data extracted by edge-side data processing and feature extraction units; An optimized attention mechanism model; Optimize the weight matrix of the attention mechanism model for the prediction stage, and adjust it according to the historical power quality status data and the prediction target; Optimize the bias vector of the attention mechanism model for the prediction stage; For edge computing models; To predict the parameters of the edge computing model in the prediction stage, the formula learns from historical power quality parameters such as voltage sag and three-phase imbalance, combines them with current characteristic data, and predicts the power quality status of the transmission and distribution system in the future.

[0043] Specifically, the power quality status prediction model combines the advantages of edge computing and an optimized attention mechanism. By learning from historical data and analyzing current features, it accurately predicts future power quality conditions. When predicting voltage sag events, the model comprehensively considers multiple factors, such as system load changes and equipment operating status, providing early warning of potential voltage sag risks and predicting their duration and impact. This predictive capability enables the system to take preventive measures, such as adjusting the output of reactive power compensation devices, effectively mitigating the impact of voltage sags on sensitive equipment.

[0044] Preferably, the formula for generating the control strategy by the dynamic control strategy generation unit is: ,in, Represents the generated dynamic control strategy; Power quality status data predicted by the power quality status assessment and prediction unit; Represents the operating constraints of the transmission and distribution system, including equipment capacity limitations, safe operating range, etc. A policy generation model is built based on an edge computing model and an optimized attention mechanism model; The parameters of the strategy generation model are obtained by learning historical control strategies and system responses under different power quality states, and are used to generate control strategies based on the predicted power quality state and system operation constraints.

[0045] Specifically, the dynamic control strategy generation model combines prediction results with system operating constraints, enabling intelligent generation of control strategies. When addressing three-phase imbalance, the model generates an optimal control plan based on the predicted imbalance trend while taking into account constraints such as transformer capacity and line current capacity. For example, by adjusting the tap position of the distribution transformer or switching capacitor banks, the three-phase load can be balanced, thereby reducing line losses and improving equipment utilization.

[0046] Preferably, after sending the control instruction to the control device, the control instruction execution and feedback unit adopts an edge computing model for feedback processing of the execution result, and the formula is: ,in, Indicates the feedback data of the executed result after processing; Indicates the actual execution information of the control instruction executed by the device, including execution time, execution status, etc. For edge computing models; The parameters of the edge computing model in this processing process are obtained by learning historical execution results and system status, and are used to extract valuable feedback data for power quality status assessment from device execution information.

[0047] Specifically, the feedback processing model analyzes device execution information to extract valuable feedback data for power quality assessment. When evaluating the control effectiveness of reactive power compensation devices, the model analyzes parameters such as execution time and voltage change to determine whether the compensation strategy achieved its intended goal. If the actual results deviate from expectations, the system adjusts subsequent control strategies based on the feedback data, forming a closed-loop control system and improving the system's control accuracy and stability.

[0048] Preferably, when the multi-dimensional heterogeneous data acquisition and access unit collects data, an edge computing model is used for time synchronization processing of different types of data, and the formula is: ,in, Indicates the timestamp after time synchronization; Indicates the original timestamp when the data was collected; For edge computing models; It is a time synchronization parameter determined based on factors such as clock deviation and communication delay of data acquisition equipment in the power transmission and distribution system. It is used for precise time synchronization of multi-dimensional heterogeneous data such as voltage, current, and frequency to ensure the accuracy of subsequent data processing and analysis.

[0049] Specifically, the data time synchronization model addresses clock deviation issues during multi-dimensional, heterogeneous data collection, achieving precise time synchronization through edge computing. In a distributed monitoring system, clocks at different monitoring points may exhibit slight deviations, which can affect the accurate analysis of power quality parameters. This model estimates and compensates for clock deviations and communication delays at each collection device in real time, ensuring a unified time base for all collected data, thereby improving the accuracy of subsequent data analysis and processing.

[0050] Preferably, in the optimized attention mechanism feature analysis unit, the optimized attention mechanism model introduces an adaptive weight adjustment formula when processing feature data: ,in, Indicates the The attention weight of feature data; Indicates the Characteristic data; 、 The query vector and key vector are obtained by training the power quality feature data of the transmission and distribution system; Score is a function that calculates the similarity between the feature data and the query vector and key vector. This formula analyzes the correlation between feature data such as voltage harmonic content and three-phase imbalance, adaptively adjusts the attention weights of different feature data, and enhances the ability to extract key features.

[0051] Specifically, the adaptive weight adjustment mechanism further optimizes the attention model by dynamically adjusting attention weights by calculating the similarity between feature data. When analyzing the relationship between voltage fluctuations and load changes, the model automatically identifies the degree of correlation between the two based on real-time data, assigning higher weights to features with high correlation. This adaptive adjustment capability enables the model to better adapt to power quality monitoring needs under different operating conditions, improving the system's robustness and adaptability.

[0052] Preferably, the power quality status assessment and prediction unit uses the following assessment model formula when performing status assessment: ,in, Indicates the power quality status assessment result; is the number of features involved in the evaluation; For the The weight of each feature in the evaluation is obtained by learning historical power quality data through an optimized attention mechanism model; Indicates the The formula combines multiple power quality characteristic data such as voltage amplitude, frequency deviation, and harmonic distortion rate to quantitatively evaluate the power quality status of the transmission and distribution system.

[0053] Specifically, the power quality status assessment model comprehensively considers multiple characteristic data and their weightings to achieve a quantitative assessment of power quality. When assessing the degree of harmonic pollution, the model simultaneously considers multiple indicators, such as harmonic content and harmonic distortion, and assigns different weights based on their impact on power quality. This comprehensive assessment method comprehensively and objectively reflects the power quality status and provides a scientific basis for hierarchical power quality management and improvement.

[0054] like Figure 2 As shown, a method for real-time monitoring and dynamic control of power quality in a power transmission and distribution system includes the following steps: S1. The multi-dimensional heterogeneous data acquisition and access unit collects multi-dimensional heterogeneous data such as voltage amplitude, frequency, three-phase imbalance, and harmonic content from various power quality monitoring points distributed at different locations in the power transmission and distribution system according to a predetermined sampling frequency and communication protocol, and transmits the collected data to the edge-side data processing and feature extraction unit. S2, the edge-side data processing and feature extraction unit receives the transmitted data, uses the edge computing model to perform preliminary processing on the data, such as format conversion and outlier detection, and extracts feature data closely related to power quality from the processed data based on pre-set feature extraction rules, and passes it to the optimized attention mechanism feature analysis unit; S3. After the optimized attention mechanism feature analysis unit obtains the feature data, it uses the optimized attention mechanism model to conduct an in-depth analysis of the feature data. By calculating the correlation between the feature data, it highlights the feature information that is important for power quality assessment and prediction, and sends the analyzed feature data to the power quality status assessment and prediction unit. S4. The power quality status assessment and prediction unit uses the edge computing model and optimized attention mechanism model based on the received feature data, combined with historical power quality data and system operating parameters, to assess the power quality status of the current transmission and distribution system and predict the power quality status in the future. The assessment and prediction results are transmitted to the dynamic control strategy generation unit. S5. The dynamic control strategy generation unit generates a dynamic control strategy for power quality issues based on the received power quality status assessment and prediction results and the current operating constraints of the transmission and distribution system using a strategy generation algorithm based on an edge computing model and an optimized attention mechanism model, and sends the control strategy to the control instruction execution and feedback unit. S6. The control instruction execution and feedback unit receives the control strategy, converts the control instruction into a control signal suitable for execution by the control equipment of the power transmission and distribution system, and sends it. It monitors the execution status in real time during the process of the equipment executing the control instruction. After the execution is completed, the execution result is processed and fed back to the power quality status assessment and prediction unit for the next round of power quality status assessment and control strategy generation.

[0055] A real-time monitoring and dynamic control system for power quality in power transmission and distribution systems, through innovative architecture design and model application, fundamentally solves the two major challenges of traditional technologies: low data processing efficiency and insufficient analysis and control accuracy. In terms of data processing, the system shifts computing tasks to the edge. After the multi-dimensional heterogeneous data acquisition and access unit acquires power quality parameters such as voltage amplitude and frequency, the edge-side data processing and feature extraction unit directly completes data format conversion, outlier removal, and feature extraction locally. This avoids the network congestion and processing delays caused by uploading all raw data to the cloud, greatly improving data processing speed and enabling the system to quickly respond to power quality issues.

[0056] In terms of analyzing and controlling precision, the system leverages an optimized attention mechanism model to deeply explore the complex connections between power quality parameters. The optimized attention mechanism feature analysis unit performs in-depth analysis of the extracted feature data, identifying key features for power quality assessment and prediction. The power quality status assessment and prediction unit combines an edge computing model with the optimized attention mechanism model, integrating historical data with current features to accurately assess and predict power quality status. The dynamic control strategy generation unit then generates highly targeted control strategies based on the assessment and prediction results and system operating constraints. The control instruction execution and feedback unit ensures the precise implementation of the strategy and provides timely feedback on the execution results, forming a complete closed loop and achieving high-precision dynamic control of power quality.

[0057] The system's advantages also lie in its intelligent collaboration throughout the entire process. From data collection, processing, and analysis to state assessment, prediction, control strategy generation, and instruction execution feedback, each link works closely together. The multi-dimensional, heterogeneous data collection and access unit ensures comprehensive and accurate data. Subsequent units leverage edge computing models and optimized attention mechanism models to achieve efficient data processing and deep mining. The control instruction execution and feedback unit ensures continuous system optimization. This intelligent collaboration throughout the entire process significantly improves the efficiency and reliability of power quality monitoring and control within the transmission and distribution system, providing a solid foundation for the stable operation of the smart grid.

[0058] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," "connected," and "fixed" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; they may refer to mechanical connections or electrical connections; they may refer to direct connections or indirect connections through an intermediate medium; and they may refer to internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0059] While embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various equivalent changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A real-time monitoring and dynamic control system for power quality of power transmission and distribution system, characterized in that: include: Multi-dimensional heterogeneous data acquisition and access unit: used to collect multi-dimensional heterogeneous data from the power transmission and distribution system, including voltage amplitude, frequency, three-phase imbalance, harmonic content, voltage sag and swell, and power quality parameters, and access the collected data into the system through an adapted communication protocol; Edge-side data processing and feature extraction unit: Receives data transmitted by the multi-dimensional heterogeneous data acquisition and access unit, performs preliminary data processing using the edge computing model, and extracts feature data related to power quality; Optimize the attention mechanism feature analysis unit: obtain the feature data output by the edge-side data processing and feature extraction unit, analyze the feature data through the optimized attention mechanism model, and mine the correlation features between the data; Power quality status assessment and prediction unit: Based on the features analyzed by the optimized attention mechanism feature analysis unit, it assesses the power quality status of the transmission and distribution system and predicts the power quality status using the edge computing model and the optimized attention mechanism model; Dynamic control strategy generation unit: Based on the results of the power quality status assessment and prediction unit and combined with the operating constraints of the transmission and distribution system, it generates a dynamic control strategy for power quality issues; Control instruction execution and feedback unit: receives the control strategy generated by the dynamic control strategy generation unit, sends the control instruction to the control equipment of the transmission and distribution system for execution, and feeds back the execution result to the power quality status assessment and prediction unit.

2. A real-time monitoring and dynamic control system for power quality of a power transmission and distribution system according to claim 1, characterized in that: The edge-side data processing and feature extraction unit uses an edge computing model to process data. The specific formula is: ,in, Represents the extracted feature data; Represents the original data collected by the multi-dimensional heterogeneous data collection and access unit; For edge computing models; It is a parameter of the edge computing model, which is obtained by training the historical power quality data of the transmission and distribution system. It is used to control the way the edge computing model processes the original data and extract characteristic data related to the voltage amplitude change trend and frequency fluctuation characteristics of the power quality from the original data.

3. A real-time monitoring and dynamic control system for power quality of a power transmission and distribution system according to claim 1, characterized in that: The optimized attention mechanism model formula adopted by the optimized attention mechanism feature analysis unit is: ,in, Represents feature data after analysis by the optimized attention mechanism; Feature data extracted by edge-side data processing and feature extraction units; The weight matrix of the attention mechanism model, whose dimension is set according to the dimension of the feature data, is used to adjust the weight of different feature data in the analysis process; is a bias vector used to enhance the expressiveness of the model, which learns the correlation between voltage, frequency, and harmonic power quality parameters in the transmission and distribution system.

4. A real-time monitoring and dynamic control system for power quality of a power transmission and distribution system according to claim 1, characterized in that: The power quality status assessment and prediction unit uses an edge computing model and an optimized attention mechanism model to predict the power quality status. The formula is: ,in, Indicates predicted power quality status data; Feature data extracted by edge-side data processing and feature extraction units; An optimized attention mechanism model; Optimize the weight matrix of the attention mechanism model for the prediction stage, and adjust it according to the historical power quality status data and the prediction target; Optimize the bias vector of the attention mechanism model for the prediction stage; For edge computing models; To predict the parameters of the edge computing model in the prediction stage, the formula predicts the power quality status of the transmission and distribution system in the future by learning the historical voltage sag and three-phase imbalance power quality parameters and combining them with the current characteristic data.

5. The real-time monitoring and dynamic control system for power quality of a power transmission and distribution system according to claim 1 is characterized in that: The formula for generating the control strategy by the dynamic control strategy generation unit is: ,in, Represents the generated dynamic control strategy; Power quality status data predicted by the power quality status assessment and prediction unit; Represents the operating constraints of the transmission and distribution system, including equipment capacity limitations and safe operating range; A policy generation model is built based on an edge computing model and an optimized attention mechanism model; The parameters of the strategy generation model are obtained by learning historical control strategies and system responses under different power quality states, and are used to generate control strategies based on the predicted power quality state and system operation constraints.

6. A real-time monitoring and dynamic control system for power quality of a power transmission and distribution system according to claim 1, characterized in that: After sending the control instruction to the control device, the control instruction execution and feedback unit uses the edge computing model to process the feedback of the execution result. The formula is: ,in, Indicates the feedback data of the executed result after processing; Indicates the actual execution information of the control instruction executed by the device, including execution time and execution status; For edge computing models; The parameters of the edge computing model in this processing process are obtained by learning historical execution results and system status, and are used to extract valuable feedback data for power quality status assessment from device execution information.

7. A real-time monitoring and dynamic control system for power quality of a power transmission and distribution system according to claim 1, characterized in that: When the multi-dimensional heterogeneous data acquisition and access unit collects data, the edge computing model is used for time synchronization processing of different types of data. The formula is: ,in, Indicates the timestamp after time synchronization; Indicates the original timestamp when the data was collected; For edge computing models; It is a time synchronization parameter determined based on the clock deviation and communication delay factors of the data acquisition equipment in the transmission and distribution system. It is used for accurate time synchronization of multi-dimensional heterogeneous data such as voltage, current, and frequency.

8. The real-time monitoring and dynamic control system for power quality of a power transmission and distribution system according to claim 1, characterized in that: In the optimized attention mechanism feature analysis unit, the optimized attention mechanism model introduces an adaptive weight adjustment formula when processing feature data: ,in, Indicates the The attention weight of feature data; Indicates the Characteristic data; 、 are the query vector and key vector obtained by training the power quality feature data of the transmission and distribution system; Score is a function that calculates the similarity between the feature data and the query vector and key vector. This formula adaptively adjusts the attention weights of different feature data by analyzing the correlation between the voltage harmonic content and three-phase imbalance feature data.

9. The real-time monitoring and dynamic control system for power quality of a power transmission and distribution system according to claim 1, characterized in that: When the power quality status assessment and prediction unit performs status assessment, the assessment model formula used is: ,in, Indicates the power quality status assessment result; is the number of features involved in the evaluation; For the The weight of each feature in the evaluation is obtained by learning historical power quality data through an optimized attention mechanism model; Indicates the The formula combines multiple power quality characteristic data such as voltage amplitude, frequency deviation, and harmonic distortion rate to quantitatively evaluate the power quality status of the transmission and distribution system.

10. A method for real-time monitoring and dynamic control of power quality in a power transmission and distribution system, characterized in that: The following steps are involved: S1. The multi-dimensional heterogeneous data acquisition and access unit collects multi-dimensional heterogeneous data on voltage amplitude, frequency, three-phase imbalance, and harmonic content from various power quality monitoring points distributed at different locations in the power transmission and distribution system according to a predetermined sampling frequency and communication protocol, and transmits the collected data to the edge-side data processing and feature extraction unit. S2, the edge-side data processing and feature extraction unit receives the transmitted data, uses the edge computing model to perform format conversion and outlier detection on the data, extracts feature data closely related to power quality from the processed data based on pre-set feature extraction rules, and passes it to the optimized attention mechanism feature analysis unit; S3. After the optimized attention mechanism feature analysis unit obtains the feature data, it uses the optimized attention mechanism model to conduct an in-depth analysis of the feature data. By calculating the correlation between the feature data, it highlights the feature information that is important for power quality assessment and prediction, and sends the analyzed feature data to the power quality status assessment and prediction unit. S4. The power quality status assessment and prediction unit uses the edge computing model and optimized attention mechanism model based on the received feature data, combined with historical power quality data and system operating parameters, to assess the power quality status of the current transmission and distribution system and predict the power quality status in the future. The assessment and prediction results are transmitted to the dynamic control strategy generation unit. S5. The dynamic control strategy generation unit generates a dynamic control strategy for power quality issues based on the received power quality status assessment and prediction results and the current operating constraints of the transmission and distribution system using a strategy generation algorithm based on an edge computing model and an optimized attention mechanism model, and sends the control strategy to the control instruction execution and feedback unit. S6. The control instruction execution and feedback unit receives the control strategy, converts the control instruction into a control signal suitable for execution by the control equipment of the power transmission and distribution system, and sends it. It monitors the execution status in real time during the process of the equipment executing the control instruction. After the execution is completed, the execution result is processed and fed back to the power quality status assessment and prediction unit for the next round of power quality status assessment and control strategy generation.

Images