A power quality disturbance source positioning system and method for a power distribution network

Abstract

This invention discloses a power quality disturbance source location system and method for power distribution networks, specifically relating to the field of power quality technology. It includes a holographic perception edge collaboration module, a digital twin-driven positioning engine module, a blockchain-based evidence storage and trust calculation module, and an adaptive self-healing control module. The invention employs an improved AdHoc protocol to construct a self-organizing network and utilizes lightweight AI models and sparse coding techniques to achieve real-time disturbance signal capture and data compression, improving perception efficiency. The digital twin-driven positioning engine module integrates GIS maps and power grid topology data to construct a dynamic digital twin model. Combined with multi-physics coupling analysis and traveling wave positioning algorithms, it controls the disturbance source positioning error to within 50 meters, achieving precise positioning. The blockchain-based evidence storage and trust calculation module adopts a consortium blockchain architecture and zero-knowledge proof technology to ensure the immutability of disturbance data and secure cross-entity sharing.

Description

Technical Field

[0001] This invention relates to the field of power quality technology, and in particular to a power quality disturbance source location system and method for power distribution networks. Background Technology

[0002] During the operation of power distribution networks, power quality issues are becoming increasingly prominent. Various disturbances, such as voltage dips, harmonics, and frequency fluctuations, can affect user equipment and the stability of the power grid. Existing power quality monitoring methods typically rely on centralized measurement and traditional sensor networks, which have certain limitations in terms of data real-time performance, accuracy, and comprehensiveness. The complexity of multi-physics disturbances and the dynamic changes in the power grid topology make it difficult to accurately locate the disturbance source and trace its propagation path. Existing methods are prone to problems such as data lag, excessive redundant information, and large location errors when processing large-scale, multi-source heterogeneous data.

[0003] In existing technologies, most methods rely on a single type of sensor or centralized data processing, lacking the ability to perform real-time feature extraction and anomaly screening at the source, making it difficult to capture key disturbance information in a timely manner. In terms of disturbance source localization, traditional simulation models and single algorithms cannot fully integrate multi-physics field data such as electromagnetic transients, heat conduction, and mechanical vibration, resulting in limited localization accuracy and response speed. Existing data storage and sharing mechanisms lack reliable guarantees, and cross-entity data collaboration poses security and privacy risks. In terms of control strategy execution, there is also a lack of mechanisms for rapid assessment of disturbance impacts and dynamic adjustment of load allocation, making the isolation of power quality disturbances and the system's self-healing capabilities insufficient. Summary of the Invention

[0004] The purpose of this invention is to overcome the above-mentioned shortcomings and provide a power quality disturbance source location system and method for power distribution networks, so as to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A power quality disturbance source location system for power distribution networks includes: The holographic sensing edge collaboration module is used to achieve real-time capture and preliminary processing of disturbance signals through multi-dimensional sensing technology and edge computing collaboration; The digital twin-driven positioning engine module is used to construct a digital twin of the power distribution network to achieve accurate positioning of disturbance sources and tracking of propagation paths through virtual-real mapping and dynamic simulation. The blockchain evidence storage and trust computing module is used to build a trusted data environment through blockchain technology to achieve end-to-end evidence storage and multi-entity collaborative computing of disturbed data. The adaptive and self-healing control module is used to automatically generate control strategies based on the positioning results and coordinate with various power grid devices to achieve disturbance source isolation and system self-healing.

[0006] Preferably, the holographic sensing edge collaboration module includes: The heterogeneous sensor self-organizing network module is used to deploy various types of sensors, including voltage, current, and transient signals, to achieve dynamic networking and data exchange between nodes through a self-organizing network protocol. The edge node intelligent preprocessing module is used to extract features and screen anomalies from raw data by embedding a lightweight AI model in the sensing node, and only upload key perturbation information. The wireless signal fingerprint acquisition module is used to collect the attenuation characteristics of power line carrier and LoRa wireless signals in the power distribution network to help locate the physical location of disturbance sources. An adaptive acquisition and compression module is used to automatically adjust the sampling accuracy according to power grid frequency fluctuations and to compress redundant data using sparse coding technology. The cross-media data fusion interface module is used to build a digital twin of the distribution network to achieve accurate location of disturbance sources and tracking of propagation paths through virtual-real mapping and dynamic simulation. The heterogeneous sensor self-organizing network module provides the edge node intelligent preprocessing module with raw data collected by multiple types of sensors through dynamic networking and data exchange between nodes; the edge node intelligent preprocessing module provides the wireless signal fingerprint acquisition module with key disturbance information by extracting features and filtering anomalies; the wireless signal fingerprint acquisition module provides the adaptive acquisition compression module with auxiliary positioning data by collecting attenuation features; and the adaptive acquisition compression module provides the cross-media data fusion interface module with an efficiently integrated data stream through compressed and optimized data.

[0007] Preferably, the digital twin-driven positioning engine module includes: The power grid digital twin modeling module is used to construct a dynamic digital twin model containing line parameters and load characteristics based on GIS maps and power grid topology data. The disturbance propagation dynamic simulation module is used to simulate the propagation and attenuation laws of different types of disturbances in the power grid and generate a spatiotemporal distribution map of disturbance diffusion. The multiphysics coupling analysis module is used to integrate multiphysics data such as electromagnetic transients, heat conduction, and mechanical vibration to identify composite features generated by disturbances. The reverse tracking algorithm library module is used to integrate multiple algorithm engines based on wave equations, impedance matrices, and traveling wave time differences to realize reverse location calculation of disturbance sources; The location result visualization module is used to mark the location of disturbance sources, propagation paths, and impact range in the digital twin model in real time, and present them in a three-dimensional visualization. The power grid digital twin modeling module provides the basic environment for power grid operation to the disturbance propagation dynamic simulation module by establishing a dynamic digital twin model; the disturbance propagation dynamic simulation module provides disturbance propagation reference data to the multiphysics coupling analysis module by generating a disturbance spatiotemporal distribution map; the multiphysics coupling analysis module provides the calculation basis for the reverse tracing algorithm library module by identifying the composite features of disturbances; and the reverse tracing algorithm library module provides accurate three-dimensional annotation data to the positioning result visualization module by calculating the positioning results.

[0008] Preferably, the blockchain evidence storage trust calculation module includes: The distributed data storage module is used to store key disturbance data in a chain structure on distributed nodes to ensure that the data is tamper-proof and traceable. The data credibility assessment module is used to generate credibility weight values ​​for each data source based on the historical behavior of nodes and data consistency verification. The smart contract triggering module is used to preset disturbance data sharing and location result verification smart contracts to achieve cross-entity automatic collaboration; The cross-domain data authorization interface module is used to achieve secure data sharing across departments without exposing the original data, using zero-knowledge proof technology. The historical disturbance case blockchain module is used to store the feature parameters and location results of historical disturbance events, and to build a reusable disturbance feature database. The distributed data storage module provides a reliable source of original data for the data credibility assessment module through a chained storage structure; the data credibility assessment module provides a data decision-making basis for the smart contract triggering module by generating credibility weight values; the smart contract triggering module provides executable data sharing rules for the cross-domain data authorization interface module through an automatic collaboration mechanism; and the cross-domain data authorization interface module provides storable reusable data for the historical disturbance case blockchain module through securely shared data.

[0009] Preferably, the adaptive self-healing control module includes: The rapid disturbance impact assessment module is used to calculate the degree and duration of the impact of disturbances on sensitive loads, important users, and power grid stability; The distributed collaborative control module is used to generate distributed control commands through collaborative decision-making between edge nodes and the control center. The self-healing strategy generation module is used to automatically match predefined control strategies based on the type of disturbance. The intelligent load transfer module is used to dynamically adjust load allocation based on topology analysis and transfer the load in the affected area to the backup line; A control effect feedback learning module used to record the implementation effect of control measures and optimize subsequent control strategies through reinforcement learning; The rapid disturbance impact assessment module provides decision-making basis for the distributed collaborative control module by assessing the disturbance impact; the distributed collaborative control module provides execution targets for the self-healing strategy generation module by generating control commands; the self-healing strategy generation module provides control schemes for the intelligent load transfer module by matching predefined strategies; and the intelligent load transfer module provides control execution results for the control effect feedback learning module by adjusting load allocation, which are used to optimize strategies.

[0010] In addition, this invention also discloses a method for locating power quality disturbance sources in distribution networks, applied to the aforementioned power quality disturbance source location system for distribution networks, comprising the following steps: S1. Intelligent preprocessing of multi-source heterogeneous data acquisition; S2, Multiphysics Coupled Disturbance Localization Modeling; S3, Distributed Trusted Data Processing Mechanism; S4. Adaptive control strategy generation and optimization.

[0011] Furthermore, step S1 specifically includes the following process: S1.1 Deploy multiple types of sensors, including voltage transformers (PT) and current transformers (CT), at key nodes of the distribution network. Use the improved AdHoc protocol to build a self-organizing network. Determine the optimal sampling frequency through joint analysis in the time and frequency domains. Establish a sensor node association matrix to achieve synchronous data acquisition in time and space. S1.2 Deploy a lightweight CNN-GRU fusion model at the edge nodes, use a sliding time window mechanism to monitor data anomalies, and trigger the upload of key data through a perturbation confidence formula. The perturbation confidence formula is as follows:

[0012] In the formula, Cdist(t) is the perturbation confidence level at time t, ω i Let μ be the feature weight, xi(t) be the real-time feature value, and μ be the feature weight. i σ i These represent the mean and standard deviation under normal conditions. S1.3. Acquire the received power of the power line carrier PLC and LoRa signals, and achieve coarse location of the disturbance source based on the logarithmic distance path loss model. The logarithmic distance path loss model is as follows: PL(d) = PL(d0) + 10n log 10 (d / d0)+X σ ; In the formula, PL(d) is the path loss at distance d, n is the path loss exponent, and X... σ For shadow fading, d0 is the near-range reference distance, and d is the actual distance between the transmitter and receiver; S1.4. A complete dictionary is constructed using the K-SVD algorithm, and the sparse coefficients are solved using the OMP algorithm. Key features with a non-zero coefficient ratio of less than 5% are retained, achieving a data compression ratio of more than 10:1.

[0013] Furthermore, step S2 specifically includes the following process: S2.1 Based on the CIM / E standard, the power grid topology is analyzed, GIS geographic information and real-time measurement data are integrated, a dynamic equivalent circuit model is constructed, the line is discretized using the finite element method and a node admittance matrix is ​​established to realize real-time mapping of the power grid operation status. S2.2 Build a disturbance propagation model using PSCAD / EMTDC, calculate the traveling wave propagation characteristics based on the improved Berylon model, and generate spatiotemporal distribution maps of different types of disturbances; S2.3 Collect non-electrical data such as line temperature and mechanical vibration, and adaptively adjust the multi-physics feature weights using the entropy weight method to achieve accurate identification of perturbation composite features; S2.4. Based on the principle of dual-end traveling wave positioning, the arrival time difference Δt of the traveling wave front is used to accurately locate the disturbance source through a formula combined with the impedance matrix method, with the error controlled within 50 meters. The formula is as follows:

[0014] In the formula, x represents the distance from the disturbance source to one end of the line, L is the line length, and v is the traveling wave propagation speed combined with the impedance matrix method.

[0015] Furthermore, step S3 specifically includes the following process: S3.1. A distributed storage network is built using a consortium blockchain architecture. Data digests are generated using the SHA-256 hash algorithm, and blocks are formed according to timestamps to ensure that the data is tamper-proof and traceable. S3.2 Establish a node reputation evaluation model, dynamically update the data source weight coefficients, and ensure data reliability; S3.3 Based on zero-knowledge proof ZKP technology, the Groth16 algorithm is used to optimize the proof efficiency and realize secure data sharing among power companies, users and manufacturers; S3.4 Construct a perturbation feature database and use an improved KNN algorithm for case matching to assist in localization decision-making.

[0016] Furthermore, step S4 specifically includes the following process: S4.1. Based on the Analytic Hierarchy Process (AHP), an evaluation index system is constructed, and the degree of impact on sensitive loads and important users is determined by the fuzzy comprehensive evaluation method, providing a quantitative basis for control decisions. S4.2. The Distributed Model Predictive Control (DMPC) algorithm is adopted to decompose the control problem into sub-problems and achieve global optimal control through the Alternating Direction Multiplier (ADMM) method. S4.3 Construct a load transfer optimization model with the goal of minimizing network loss. Use an improved particle swarm optimization (PSO) algorithm to solve for the optimal load allocation scheme using the following formula: v ik+1 = ωv ik + c1r1(pbest i - x ik ) + c2r2(gbest - x ik ); x ik+1 = x ik + v ik+1 ; In the formula, i represents the i-th particle, k is the current iteration number, and v ik Let x be the velocity of the i-th particle in the k-th iteration. ik Let ω be the position of the i-th particle in the k-th iteration, ω be the inertia weight, c1 and c2 be the individual learning factor and the social learning factor, respectively, and r1 and r2 be random numbers between [0,1]. pbest i Let be the individual historical best position of particle i, and gbest be the global historical best position of the population. The algorithm iteratively optimizes the velocity and position update formulas to find the optimal load allocation scheme with the goal of minimizing network loss. S4.4 Construct a deep reinforcement learning (DRL) model, with disturbance type and system state as the state space and control measures as the action space. Optimize the policy generation logic through the DQN algorithm to achieve autonomous evolution of the control policy.

[0017] Beneficial effects of this invention: 1. This invention employs an improved AdHoc protocol to construct a self-organizing network and utilizes lightweight AI models and sparse coding technology to achieve real-time capture and data compression of disturbance signals, thereby improving sensing efficiency. The digital twin-driven positioning engine module integrates GIS maps and power grid topology data to construct a dynamic digital twin model. Combined with multi-physics coupling analysis and traveling wave positioning algorithms, it controls the disturbance source positioning error within 50 meters, achieving precise positioning. The blockchain evidence storage and trust calculation module adopts a consortium blockchain architecture and zero-knowledge proof technology to ensure that disturbance data is tamper-proof and securely shared across entities. The adaptive self-healing control module generates optimization strategies through analytic hierarchy process and deep reinforcement learning, enabling rapid isolation of disturbance sources and system self-healing, thereby improving power grid stability.

[0018] 2. This invention utilizes a lightweight CNN-GRU model and sparse coding technology to achieve key data screening and compression, improving data processing efficiency; in the multi-physics coupled disturbance localization modeling stage, a dynamic simulation model is built based on the CIM / E standard and PSCAD / EMTDC, and the disturbance source is accurately located by combining the traveling wave localization principle and the impedance matrix method; in the distributed trusted data processing mechanism stage, consortium blockchain and zero-knowledge proof technology are used to ensure trusted data storage and secure sharing; in the adaptive control strategy generation and optimization stage, the load transfer and control strategies are optimized through the analytic hierarchy process and deep reinforcement learning to achieve system autonomous evolution and self-healing. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the system modules in this invention.

[0020] Figure 2 This is a schematic diagram of the method steps in this invention. Detailed Implementation

[0021] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0022] Example 1: like Figure 1 As shown, a power quality disturbance source location system for power distribution networks includes a holographic perception edge collaboration module for real-time capture and preliminary processing of disturbance signals through multi-dimensional sensing technology and edge computing; a digital twin-driven positioning engine module for constructing a digital twin of the power distribution network to achieve accurate positioning of disturbance sources and tracking of propagation paths through virtual-real mapping and dynamic simulation; a blockchain-based evidence storage and trust computing module for constructing a trusted data environment through blockchain technology to achieve full-process evidence storage of disturbance data and multi-subject collaborative computing; and an adaptive self-healing control module for automatically generating control strategies based on positioning results to coordinate various power grid equipment to achieve disturbance source isolation and system self-healing. The holographic perception edge collaboration module includes a heterogeneous sensor self-organizing network module for deploying various types of sensors such as voltage, current, and transient signals to achieve dynamic networking and data communication between nodes through a self-organizing network protocol; an edge node intelligent preprocessing module for embedding lightweight AI models in perception nodes to extract features and screen anomalies in raw data, and only uploading key disturbance information; a wireless signal fingerprint acquisition module for collecting attenuation characteristics of power line carrier and LoRa wireless signals in the distribution network to assist in locating the physical location of disturbance sources; an adaptive acquisition and compression module for automatically adjusting sampling accuracy according to power grid frequency fluctuations and compressing redundant data using sparse coding technology; and a cross-media data fusion interface module for constructing a digital twin of the distribution network to achieve accurate location of disturbance sources and tracking of propagation paths through virtual-real mapping and dynamic simulation. The heterogeneous sensor self-organizing network module provides the edge node intelligent preprocessing module with raw data collected by multiple types of sensors through dynamic networking and data exchange between nodes; the edge node intelligent preprocessing module provides the wireless signal fingerprint acquisition module with key disturbance information by extracting features and filtering anomalies; the wireless signal fingerprint acquisition module provides the adaptive acquisition compression module with auxiliary positioning data by collecting attenuation features; and the adaptive acquisition compression module provides the cross-media data fusion interface module with an efficiently integrated data stream through compressed and optimized data. The digital twin-driven positioning engine module includes a power grid digital twin modeling module for constructing a dynamic digital twin model containing line parameters and load characteristics based on GIS maps and power grid topology data; a disturbance propagation dynamic simulation module for simulating the propagation and attenuation laws of different types of disturbances in the power grid and generating a spatiotemporal distribution map of disturbance diffusion; a multi-physics coupling analysis module for integrating electromagnetic transient, heat conduction, and mechanical vibration multi-physics field data to identify the composite features generated by disturbances; a reverse tracing algorithm library module for integrating a multi-algorithm engine based on wave equations, impedance matrices, and traveling wave time differences to realize reverse positioning calculation of disturbance sources; and a positioning result visualization module for real-time annotation of the location of disturbance sources, propagation paths, and influence range in the digital twin model and three-dimensional visualization presentation. The power grid digital twin modeling module provides the basic operating environment for the disturbance propagation dynamic simulation module by establishing a dynamic digital twin model; the disturbance propagation dynamic simulation module provides disturbance propagation reference data for the multiphysics coupling analysis module by generating a disturbance spatiotemporal distribution map; the multiphysics coupling analysis module provides the calculation basis for the reverse tracing algorithm library module by identifying the complex characteristics of disturbances; and the reverse tracing algorithm library module provides accurate three-dimensional annotation data for the positioning result visualization module by calculating the positioning results. The blockchain evidence storage trust calculation module includes a distributed data storage module for storing key disturbance data in a chain structure on distributed nodes to ensure data immutability and traceability; a data credibility assessment module for generating credibility weight values ​​for each data source based on node historical behavior and data consistency verification; a smart contract triggering module for pre-setting disturbance data sharing and location result verification smart contracts to achieve cross-entity automatic collaboration; a cross-domain data authorization interface module for achieving cross-departmental data secure sharing without exposing the original data using zero-knowledge proof technology; and a historical disturbance case blockchain module for storing feature parameters and location results of historical disturbance events to construct a reusable disturbance feature database. The distributed data storage module provides a reliable source of original data for the data credibility assessment module through a chained storage structure; the data credibility assessment module provides a data decision-making basis for the smart contract triggering module by generating credibility weight values; the smart contract triggering module provides executable data sharing rules for the cross-domain data authorization interface module through an automatic collaboration mechanism; and the cross-domain data authorization interface module provides storable reusable data for the historical disturbance case blockchain module through securely shared data. The adaptive self-healing control module includes a rapid disturbance impact assessment module for calculating the degree and duration of the impact of disturbances on sensitive loads, important users, and power grid stability; a distributed collaborative control module for generating distributed control commands through collaborative decision-making between edge nodes and the control center; a self-healing strategy generation module for automatically matching predefined control strategies according to the disturbance type; a load intelligent transfer module for dynamically adjusting load allocation based on topology analysis to transfer loads in affected areas to backup lines; and a control effect feedback learning module for recording the implementation effect of control measures and optimizing the logic for generating subsequent control strategies through reinforcement learning. The rapid disturbance impact assessment module provides decision-making basis for the distributed collaborative control module by assessing the disturbance impact; the distributed collaborative control module provides execution targets for the self-healing strategy generation module by generating control commands; the self-healing strategy generation module provides control schemes for the intelligent load transfer module by matching predefined strategies; and the intelligent load transfer module provides control execution results for the control effect feedback learning module by adjusting load allocation, which are used to optimize strategies.

[0023] In the above embodiments, the holographic perception edge collaboration module deploys multiple types of sensors such as voltage and current, constructs a self-organizing network using an improved AdHoc protocol, and achieves real-time capture and data compression of disturbance signals through a lightweight AI model and sparse coding technology, thereby improving perception efficiency. The digital twin-driven positioning engine module integrates GIS maps and power grid topology data to construct a dynamic digital twin model. Combining multi-physics coupling analysis and traveling wave positioning algorithms, it controls the disturbance source positioning error within 50 meters, achieving accurate positioning. The blockchain evidence storage and trust calculation module adopts a consortium blockchain architecture and zero-knowledge proof technology to ensure that disturbance data is tamper-proof and securely shared across entities. The adaptive self-healing control module generates optimization strategies through analytic hierarchy process and deep reinforcement learning, achieving rapid isolation of disturbance sources and system self-healing, thereby improving power grid stability.

[0024] Example 2: like Figure 2 As shown, a method for locating power quality disturbance sources in a distribution network includes the following steps: S1. Intelligent preprocessing of multi-source heterogeneous data acquisition; S2, Multiphysics Coupled Disturbance Localization Modeling; S3, Distributed Trusted Data Processing Mechanism; S4. Adaptive control strategy generation and optimization; In step S1, intelligent preprocessing of multi-source heterogeneous data acquisition: S1.1 Deploy multiple types of sensors, such as voltage transformers (PT) and current transformers (CT), at key nodes of the distribution network. Use the improved AdHoc protocol to build a self-organizing network. Determine the optimal sampling frequency through joint analysis in the time and frequency domains. Establish a sensing node association matrix to achieve spatiotemporal synchronous data acquisition. S1.2 Deploy a lightweight CNN-GRU fusion model at the edge nodes, use a sliding time window mechanism to monitor data anomalies, and trigger the upload of key data through a perturbation confidence formula, as follows:

[0025] In the formula, Cdist(t) is the perturbation confidence level at time t, ω i Let μ be the feature weight, xi(t) be the real-time feature value, and μ be the feature weight. i σ i These represent the mean and standard deviation under normal conditions. S1.3. Acquire the received power of the power line carrier PLC and LoRa signals, and achieve coarse location of the disturbance source based on the logarithmic distance path loss model. The logarithmic distance path loss model is as follows: PL(d) = PL(d0) + 10n log 10 (d / d0)+X σ ; In the formula, PL(d) is the path loss at distance d, n is the path loss exponent, and X... σ For shadow fading, d0 is the near-range reference distance, and d is the actual distance between the transmitter and receiver; S1.4. The K-SVD algorithm is used to construct an overcomplete dictionary, and the OMP algorithm is used to solve the sparse coefficients. Key features with a non-zero coefficient ratio of less than 5% are retained to achieve a data compression ratio of more than 10:1. In step S2, multiphysics field coupled perturbation localization modeling: S2.1 Based on the CIM / E standard, the power grid topology is analyzed, GIS geographic information and real-time measurement data are integrated, a dynamic equivalent circuit model is constructed, the line is discretized using the finite element method and a node admittance matrix is ​​established to realize real-time mapping of the power grid operation status. S2.2 Build a disturbance propagation model using PSCAD / EMTDC, calculate the traveling wave propagation characteristics based on the improved Berylon model, and generate spatiotemporal distribution maps of different types of disturbances (voltage sag, harmonics); S2.3 Collect non-electrical data such as line temperature and mechanical vibration, and adaptively adjust the multi-physics feature weights using the entropy weight method to achieve accurate identification of disturbance composite features; S2.4. Based on the principle of dual-end traveling wave positioning, the arrival time difference Δt of the traveling wave front is used to accurately locate the disturbance source through a formula combined with the impedance matrix method, with the error controlled within 50 meters. The formula is as follows:

[0026] In the formula, x represents the distance from the disturbance source (fault point) to one end of the line, L is the line length, and v is the traveling wave propagation speed. In step S3, the distributed trusted data processing mechanism: S3.1. A distributed storage network is built using a consortium blockchain architecture. Data digests are generated using the SHA-256 hash algorithm, and blocks are formed according to timestamps to ensure that the data is tamper-proof and traceable. S3.2 Establish a node reputation evaluation model, dynamically update the data source weight coefficients, and ensure data reliability; S3.3 Based on zero-knowledge proof (ZKP) technology, the Groth16 algorithm is used to optimize the proof efficiency and realize secure data sharing among power companies, users, and manufacturers; S3.4 Construct a perturbation feature database and use an improved KNN algorithm for case matching to assist in localization decision-making; In step S4, during the adaptive control strategy generation and optimization: S4.1. Construct an evaluation index system based on the Analytic Hierarchy Process (AHP), and determine the degree of impact on sensitive loads and important users through fuzzy comprehensive evaluation method to provide a quantitative basis for control decisions; S4.2. The Distributed Model Predictive Control (DMPC) algorithm is adopted to decompose the control problem into sub-problems and achieve global optimal control through the Alternating Direction Multiplier Method (ADMM). S4.3 Construct a load transfer optimization model with the goal of minimizing network loss. Use an improved particle swarm optimization (PSO) algorithm to solve for the optimal load allocation scheme using the following formula: v ik+1 = ωv ik + c1r1(pbest i - x ik ) + c2r2(gbest - x ik ); x ik+1 = x ik + v ik+1 ; In the formula, i represents the i-th particle, k is the current iteration number, and v ikLet x be the velocity of the i-th particle in the k-th iteration. ik Let ω be the position of the i-th particle in the k-th iteration, ω be the inertia weight, c1 and c2 be the individual learning factor and the social learning factor, respectively, and r1 and r2 be random numbers between [0,1]. pbest i Let be the individual historical best position of particle i, and gbest be the global historical best position of the population. The algorithm iteratively optimizes the velocity and position update formulas to find the optimal load allocation scheme with the goal of minimizing network loss. S4.4 Construct a deep reinforcement learning (DRL) model, with the disturbance type and system state as the state space and the control measures as the action space. Optimize the policy generation logic through the DQN algorithm to achieve autonomous evolution of the control policy.

[0027] In the above embodiments, during the intelligent preprocessing stage of multi-source heterogeneous data acquisition, multiple types of sensors are deployed to build a self-organizing network. A lightweight CNN-GRU model and sparse coding technology are used to achieve key data screening and compression, improving data processing efficiency. In the multi-physics coupled disturbance localization modeling stage, a dynamic simulation model is built based on the CIM / E standard and PSCAD / EMTDC. The traveling wave localization principle and impedance matrix method are combined to achieve accurate localization of disturbance sources. In the distributed trusted data processing mechanism stage, consortium blockchain and zero-knowledge proof technology are used to ensure trusted data storage and secure sharing. In the adaptive control strategy generation and optimization stage, the load transfer and control strategies are optimized through the analytic hierarchy process and deep reinforcement learning to achieve system autonomous evolution and self-healing.

[0028] Working Principle: In use, this invention deploys multiple types of sensors, such as voltage and current sensors, through a holographic sensing edge collaboration module. It uses an improved AdHoc protocol to build a self-organizing network to achieve real-time data acquisition and transmission between nodes. Edge computing capabilities enable preliminary data processing at the source. Lightweight AI models are used to extract features and filter anomalies from signals, uploading only critical disturbance data, thus improving sensing efficiency. The attenuation characteristics of wireless signals, such as power line carrier (PLC) and LoRa signals, are used to assist in locating the disturbance source. The system compresses data using sparse coding technology to compress redundant data, thereby improving the system's response speed and processing capabilities. All of this data is integrated through a cross-media data fusion interface module, thereby achieving accurate location of the disturbance source and tracking of its propagation path.

[0029] The digital twin-driven positioning engine module integrates GIS maps and power grid topology data to construct a dynamic digital twin model of the power grid. This model can simulate the state of the power grid in real time and track the propagation patterns of different types of disturbances in the power grid, generating a spatiotemporal distribution map. The system also incorporates multi-physics data, such as electromagnetic transients, heat conduction, and mechanical vibration. Through complex multi-physics coupling analysis, it identifies the composite characteristics of disturbance sources. Using traveling wave positioning technology and impedance matrix method, the system can accurately locate disturbance sources with a control error within 50 meters. The positioning results are then reverse-calculated using a reverse tracing algorithm to accurately identify the location of the disturbance source and its propagation path. The results are then presented to the control center through a 3D visualization module for real-time monitoring.

[0030] With the support of the blockchain-based evidence storage and trust computing module, all critical disturbance data is stored in a chain structure on distributed nodes, ensuring that the data is tamper-proof and traceable. By using a consortium blockchain architecture and zero-knowledge proof technology, the system can achieve secure data sharing and protect the privacy and security of all parties. To enhance data credibility, the system dynamically updates the weight of data sources and uses smart contracts to automatically conduct cross-departmental collaboration. The adaptive self-healing control module generates a solution based on the disturbance impact assessment, isolates the disturbance source, and dynamically adjusts the load distribution according to the actual situation to ensure the stability and self-healing capability of the power grid. It effectively locates the power quality disturbance source, automatically optimizes the control strategy, and ensures the efficient and stable operation of the distribution network.

[0031] Through the multi-source heterogeneous data acquisition and intelligent preprocessing stage, multiple sensors are deployed at key nodes. Time-frequency domain joint analysis and an improved AdHoc protocol ensure spatiotemporal synchronization and accurate data acquisition. Data is intelligently analyzed and anomaly detected using a CNN-GRU fusion model. Sparse coding technology is used to achieve efficient data compression, further enhancing real-time processing capabilities. Moving into the disturbance location modeling stage, the system analyzes the power grid topology using the CIM / E standard, combines GIS geographic information with real-time measurement data, and constructs a dynamic power grid simulation model using tools such as the finite element method and PSCAD / EMTDC. Traveling wave positioning principle and dual-end positioning technology are used to accurately calculate the location of disturbance sources, achieving a positioning error control within 50 meters. The distributed trusted data processing mechanism adopts a consortium blockchain architecture, combined with zero-knowledge proof technology, to ensure secure data sharing among various entities, improving data reliability and transparency. In the adaptive control strategy generation and optimization stage, the system uses techniques such as the analytic hierarchy process (AHP) and deep reinforcement learning to optimize load transfer and system control strategies, achieving rapid isolation of disturbance sources and the power grid's self-healing capability, ensuring power quality stability and the power grid's self-repair capability.

[0032] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A power quality disturbance source location system for power distribution networks, characterized in that, include: The holographic sensing edge collaboration module is used to achieve real-time capture and preliminary processing of disturbance signals through multi-dimensional sensing technology and edge computing collaboration; The digital twin-driven positioning engine module is used to construct a digital twin of the power distribution network to achieve accurate positioning of disturbance sources and tracking of propagation paths through virtual-real mapping and dynamic simulation. The blockchain evidence storage and trust computing module is used to build a trusted data environment through blockchain technology to achieve end-to-end evidence storage and multi-entity collaborative computing of disturbed data. The adaptive and self-healing control module is used to automatically generate control strategies based on the positioning results and coordinate with various power grid devices to achieve disturbance source isolation and system self-healing.

2. The power quality disturbance source location system for distribution networks according to claim 1, characterized in that, The holographic sensing edge collaboration module includes: The heterogeneous sensor self-organizing network module is used to deploy various types of sensors, including voltage, current, and transient signals, to achieve dynamic networking and data exchange between nodes through a self-organizing network protocol. The edge node intelligent preprocessing module is used to extract features and screen anomalies from raw data by embedding a lightweight AI model in the sensing node, and only upload key perturbation information. The wireless signal fingerprint acquisition module is used to collect the attenuation characteristics of power line carrier and LoRa wireless signals in the power distribution network to help locate the physical location of disturbance sources. An adaptive acquisition and compression module is used to automatically adjust the sampling accuracy according to power grid frequency fluctuations and to compress redundant data using sparse coding technology. The cross-media data fusion interface module is used to build a digital twin of the distribution network to achieve accurate location of disturbance sources and tracking of propagation paths through virtual-real mapping and dynamic simulation. The heterogeneous sensor self-organizing network module provides the edge node intelligent preprocessing module with raw data collected by multiple types of sensors through dynamic networking and data exchange between nodes; the edge node intelligent preprocessing module provides the wireless signal fingerprint acquisition module with key disturbance information by extracting features and filtering anomalies; the wireless signal fingerprint acquisition module provides the adaptive acquisition compression module with auxiliary positioning data by collecting attenuation features; and the adaptive acquisition compression module provides the cross-media data fusion interface module with an efficiently integrated data stream through compressed and optimized data.

3. A power quality disturbance source location system for distribution networks according to claim 1, characterized in that, The digital twin-driven positioning engine module includes: The power grid digital twin modeling module is used to construct a dynamic digital twin model containing line parameters and load characteristics based on GIS maps and power grid topology data. The disturbance propagation dynamic simulation module is used to simulate the propagation and attenuation laws of different types of disturbances in the power grid and generate a spatiotemporal distribution map of disturbance diffusion. The multiphysics coupling analysis module is used to integrate multiphysics data such as electromagnetic transients, heat conduction, and mechanical vibration to identify composite features generated by disturbances. The reverse tracking algorithm library module is used to integrate multiple algorithm engines based on wave equations, impedance matrices, and traveling wave time differences to realize reverse location calculation of disturbance sources; The location result visualization module is used to mark the location of disturbance sources, propagation paths, and impact range in the digital twin model in real time, and present them in a three-dimensional visualization. The power grid digital twin modeling module provides the basic environment for power grid operation to the disturbance propagation dynamic simulation module by establishing a dynamic digital twin model; the disturbance propagation dynamic simulation module provides disturbance propagation reference data to the multiphysics coupling analysis module by generating a disturbance spatiotemporal distribution map; the multiphysics coupling analysis module provides the calculation basis for the reverse tracing algorithm library module by identifying the composite features of disturbances; and the reverse tracing algorithm library module provides accurate three-dimensional annotation data to the positioning result visualization module by calculating the positioning results.

4. A power quality disturbance source location system for distribution networks according to claim 1, characterized in that, The blockchain evidence storage trust calculation module includes: The distributed data storage module is used to store key disturbance data in a chain structure on distributed nodes to ensure that the data is tamper-proof and traceable. The data credibility assessment module is used to generate credibility weight values ​​for each data source based on the historical behavior of nodes and data consistency verification. The smart contract triggering module is used to preset disturbance data sharing and location result verification smart contracts to achieve cross-entity automatic collaboration; The cross-domain data authorization interface module is used to achieve secure data sharing across departments without exposing the original data, using zero-knowledge proof technology. The historical disturbance case blockchain module is used to store the feature parameters and location results of historical disturbance events, and to build a reusable disturbance feature database. The distributed data storage module provides a reliable source of original data for the data credibility assessment module through a chained storage structure; the data credibility assessment module provides a data decision-making basis for the smart contract triggering module by generating credibility weight values; the smart contract triggering module provides executable data sharing rules for the cross-domain data authorization interface module through an automatic collaboration mechanism; and the cross-domain data authorization interface module provides storable reusable data for the historical disturbance case blockchain module through securely shared data.

5. A power quality disturbance source location system for distribution networks according to claim 1, characterized in that, The adaptive self-healing control module includes: The rapid disturbance impact assessment module is used to calculate the degree and duration of the impact of disturbances on sensitive loads, important users, and power grid stability; The distributed collaborative control module is used to generate distributed control commands through collaborative decision-making between edge nodes and the control center. The self-healing strategy generation module is used to automatically match predefined control strategies based on the type of disturbance. The intelligent load transfer module is used to dynamically adjust load allocation based on topology analysis and transfer the load in the affected area to the backup line; A control effect feedback learning module used to record the implementation effect of control measures and optimize subsequent control strategies through reinforcement learning; The rapid disturbance impact assessment module provides decision-making basis for the distributed collaborative control module by assessing the disturbance impact; the distributed collaborative control module provides execution targets for the self-healing strategy generation module by generating control commands; the self-healing strategy generation module provides control schemes for the intelligent load transfer module by matching predefined strategies; and the intelligent load transfer module provides control execution results for the control effect feedback learning module by adjusting load allocation, which are used to optimize strategies.

6. A method for locating power quality disturbance sources in a distribution network, applied to the power quality disturbance source location system for a distribution network as described in any one of claims 1 to 5, characterized in that, Includes the following steps: S1. Intelligent preprocessing of multi-source heterogeneous data acquisition; S2, Multiphysics Coupled Disturbance Localization Modeling; S3, Distributed Trusted Data Processing Mechanism; S4. Adaptive control strategy generation and optimization.

7. The method for locating power quality disturbance sources in a distribution network according to claim 6, characterized in that: The specific steps S1 are as follows The process includes the following: S1.1 Deploy multiple types of sensors, including voltage transformers (PT) and current transformers (CT), at key nodes of the distribution network. Use the improved AdHoc protocol to build a self-organizing network. Determine the optimal sampling frequency through joint analysis in the time and frequency domains. Establish a sensor node association matrix to achieve synchronous data acquisition in time and space. S1.2 Deploy a lightweight CNN-GRU fusion model at the edge nodes, use a sliding time window mechanism to monitor data anomalies, and trigger the upload of key data through a perturbation confidence formula. The perturbation confidence formula is as follows: In the formula, Cdist(t) is the disturbance confidence at time t, ω i is the characteristic weight, xi(t) is the real-time characteristic value, μ i , σ i are the mean and standard deviation under normal state; S1.

3. Acquire the received power of the power line carrier PLC and LoRa signals, and achieve coarse location of the disturbance source based on the logarithmic distance path loss model. The logarithmic distance path loss model is as follows: PL(d) = PL(d0) + 10n log 10 (d / d0) + X σ ; where PL(d) is the path loss at distance d, n is the path loss exponent, X σ is the shadow fading, d0is the near-field reference distance, and d is the actual distance between the transmitter and receiver. S1.

4. A complete dictionary is constructed using the K-SVD algorithm, and the sparse coefficients are solved using the OMP algorithm. Key features with a non-zero coefficient ratio of less than 5% are retained, achieving a data compression ratio of more than 10:

1.

8. The method for locating power quality disturbance sources in a distribution network according to claim 6, characterized in that: Step S2 specifically includes the following process: S2.1 Based on the CIM / E standard, the power grid topology is analyzed, GIS geographic information and real-time measurement data are integrated, a dynamic equivalent circuit model is constructed, the line is discretized using the finite element method and a node admittance matrix is ​​established to realize real-time mapping of the power grid operation status. S2.2 Build a disturbance propagation model using PSCAD / EMTDC, calculate the traveling wave propagation characteristics based on the improved Berylon model, and generate spatiotemporal distribution maps of different types of disturbances; S2.3 Collect non-electrical data such as line temperature and mechanical vibration, and adaptively adjust the multi-physics feature weights using the entropy weight method to achieve accurate identification of perturbation composite features; S2.

4. Based on the principle of dual-end traveling wave positioning, the arrival time difference Δt of the traveling wave front is used to accurately locate the disturbance source through a formula combined with the impedance matrix method, with the error controlled within 50 meters. The formula is as follows: In the formula, x represents the distance from the disturbance source to one end of the line, L is the line length, and v is the traveling wave propagation speed combined with the impedance matrix method.

9. A method for locating power quality disturbance sources in a distribution network according to claim 6, characterized in that: The specific steps S3 are as follows The process includes the following: S3.

1. A distributed storage network is built using a consortium blockchain architecture. Data digests are generated using the SHA-256 hash algorithm, and blocks are formed according to timestamps to ensure that the data is tamper-proof and traceable. S3.2 Establish a node reputation evaluation model, dynamically update the data source weight coefficients, and ensure data reliability; S3.3 Based on zero-knowledge proof ZKP technology, the Groth16 algorithm is used to optimize the proof efficiency and realize secure data sharing among power companies, users and manufacturers; S3.4 Construct a perturbation feature database and use an improved KNN algorithm for case matching to assist in localization decision-making.

10. A method for locating power quality disturbance sources in a distribution network according to claim 6, characterized in that: Step S4 specifically includes the following process: S4.

1. Based on the Analytic Hierarchy Process (AHP), an evaluation index system is constructed, and the degree of impact on sensitive loads and important users is determined by the fuzzy comprehensive evaluation method, providing a quantitative basis for control decisions. S4.

2. The Distributed Model Predictive Control (DMPC) algorithm is adopted to decompose the control problem into sub-problems and achieve global optimal control through the Alternating Direction Multiplier (ADMM) method. S4.3 Construct a load transfer optimization model with the goal of minimizing network loss. Use an improved particle swarm optimization (PSO) algorithm to solve for the optimal load allocation scheme using the following formula: v ik+1 = ωv ik + c1r1(pbest i - x ik ) + c2r2(gbest - x ik ); x ik+1 = x ik + v ik+1 ; In the formula, i represents the i-th particle, k is the current iteration number, and v ik Let x be the velocity of the i-th particle in the k-th iteration. ik Let ω be the position of the i-th particle in the k-th iteration, ω be the inertia weight, c1 and c2 be the individual learning factor and the social learning factor, respectively, and r1 and r2 be random numbers between [0,1]. pbest i Let be the individual historical best position of particle i, and gbest be the global historical best position of the population. The algorithm iteratively optimizes the velocity and position update formulas to find the optimal load allocation scheme with the goal of minimizing network loss. S4.4 Construct a deep reinforcement learning (DRL) model, with disturbance type and system state as the state space and control measures as the action space. Optimize the policy generation logic through the DQN algorithm to achieve autonomous evolution of the control policy.

Images