Intelligent gridding comprehensive anti-misoperation locking system for power distribution network

Abstract

The invention provides an intelligent gridding comprehensive anti-misoperation locking system for a power distribution network, and the system comprises an optimization module which achieves the hierarchical decoupling design of the system through a distributed communication middle platform, an intelligent logic analysis engine and a programmable execution control unit. The deployment module is used for modularly deploying functional components based on the micro-service architecture, reducing the system coupling degree by using an asynchronous message queue as a cross-component communication bus in combination with a domain event-driven mechanism, and completing local logic judgment at an edge computing node through an FPGA (Field Programmable Gate Array); the network reinforcement module is used for constructing a heterogeneous communication network and guaranteeing the communication quality through a dual-channel redundancy transmission technology; the intelligent decision-making module is used for forming a self-adaptive decision-making model based on reinforcement learning and a knowledge graph technology; and the digital twinning verification module is used for establishing a power grid virtual mapping system through a digital twinning technology. Cooperation among modules of the system can be balanced, instruction processing time delay is shortened, communication reliability is improved, and efficient implementation of an operation scheme is guaranteed.

Description

Technical Field

[0001] This invention relates to the fields of smart grid and new power system technology, specifically to a smart grid-based integrated anti-misoperation interlocking system for distribution networks, applicable to multi-energy collaborative self-healing control and regional anti-misoperation scenarios in new power systems. Background Technology

[0002] The efficient operation of smart grids cannot be separated from the innovative integration of technologies. The combination of edge computing and digital twins provides a new direction for the development of power grids. This model reduces the central processing pressure through distributed computing, while accurately restoring the operating status with digital twins, thus building a more flexible and reliable system. In practical applications, optimization is needed from multiple aspects. In terms of architecture design, the functions of each module need to be rationally planned to reduce coupling and improve system flexibility. The deployment of edge nodes should comprehensively consider data distribution and computing needs to ensure real-time processing effects. In terms of communication security, encryption and authentication mechanisms need to be strengthened to ensure reliable data transmission. The construction of intelligent decision-making systems should integrate multi-source data to improve the scientific nature of regulation. The digital twin verification mechanism needs to be continuously improved to ensure the accuracy of simulation results and provide strong support for solution formulation.

[0003] Currently, this model still faces many challenges. How to balance the collaboration between various modules of the system, shorten the instruction processing latency, improve communication reliability, and ensure the efficient implementation of operation plans are urgent problems to be solved. Through technological innovation and practical exploration, we need to continuously optimize the system architecture and promote the development of smart grids to a higher level.

[0004] Therefore, how to balance the collaboration between various modules of the system, shorten the instruction processing latency, improve communication reliability, and ensure the efficient implementation of operation plans are urgent problems to be solved. We need to continuously optimize the system architecture through technological innovation and practical exploration to promote the development of smart grids to a higher level. Summary of the Invention

[0005] To address the various shortcomings of existing technologies, this invention provides a smart grid-based integrated anti-misoperation interlocking system for distribution networks with low coupling, high real-time performance, and strong fault tolerance, thereby balancing module collaboration efficiency and communication reliability.

[0006] The present invention achieves the above objectives through the following technical solutions: A smart grid-based integrated anti-misoperation interlocking system for power distribution networks includes: The optimization module is used to implement a layered decoupling design of the system through a distributed communication platform, an intelligent logic analysis engine, and a programmable execution control unit. The layered decoupling relationship of the system is represented as follows:

[0007] in, ddecouple <1, C sys For the overall coupling of the system, C modulei Let i be the independent coupling degree of the i-th module. d decouple These are the layered decoupling coefficients; The deployment module is used to modularly deploy functional components based on a microservice architecture. It utilizes an asynchronous message queue as a cross-component communication bus, combines a domain event-driven mechanism to reduce system coupling, and completes local logic judgments on edge computing nodes through FPGA programmable logic devices. The network hardening module is used to build a heterogeneous communication network of 5G and fiber optics, and ensures communication quality through dual-channel redundant transmission technology. The intelligent decision-making module is used to form an adaptive decision-making model based on reinforcement learning and knowledge graph technology, and output a standardized operation sequence. O ={ o 1, o 2,…, o m}, represented as:

[0008] in, r t Let be the reward value at time t. c As a discount factor, V ( s t+1 ) is the state s t+1 The value function; The digital twin verification module is used to establish a virtual image system of the power grid through digital twin technology and to verify the feasibility of the operation scheme using Monte Carlo simulation.

[0009] According to the present invention, a smart grid-based integrated anti-misoperation interlocking system for power distribution networks is based on a microservice architecture to implement modular design of functional components, and completes cross-component information interaction through asynchronous message queues and domain event-driven mechanisms. Specifically, it includes the following steps: Functional module division steps: The power system control function is divided into multiple independent small microservice units. Each microservice unit is responsible for a specific function or business logic, including at least load monitoring, power balancing, equipment status monitoring, and fault early warning. Each microservice unit achieves interoperability through a lightweight interface REST API. Asynchronous message queue communication steps: An asynchronous message queue is used as a cross-component communication bus. Each microservice unit publishes or subscribes to events through the message queue to achieve loosely coupled communication. The message queue supports the transmission of multiple types of events, including at least status changes, control commands, and warning information. Domain event-driven mechanism: Define a domain event model, including at least event type, event source, event data, and triggering conditions; when a microservice unit detects a specific state change, it generates a corresponding domain event and publishes it to the message queue.

[0010] According to the present invention, a smart grid-based integrated anti-misoperation interlocking system for power distribution networks completes local logic judgment at edge computing nodes through FPGA programmable logic devices, specifically including the following steps: Integrate FPGA programmable logic devices in edge computing nodes and configure them as hardware acceleration units independent of the main processor; The FPGA internally implements logic circuit partitioning, including a data acquisition module, a fault detection module, a control command generation module, and a communication interface module; The FPGA is directly connected to the power distribution feeder sensor via a high-speed ADC interface to acquire three-phase current and voltage waveform data in real time. Implement data preprocessing logic within the FPGA, including at least filtering, noise reduction, and feature extraction, to generate standardized detection parameters.

[0011] According to the intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks provided by the present invention, it also performs instantaneous identification and decision-making for emergency faults: Single-phase grounding fault detection: Parallel comparison circuits are deployed inside the FPGA to monitor the amplitude and phase of the zero-sequence current in real time; when the zero-sequence current exceeds the preset threshold, the fault flag is triggered, and the time of the fault occurrence and the feeder topology location are recorded at the same time. Short circuit fault detection: The FPGA calculates the current mutation rate through a differential algorithm. When the mutation rate exceeds the threshold, it is immediately determined to be a short circuit fault. Based on the fault type and feeder topology, the FPGA calls the corresponding action from the pre-stored strategy table. The FPGA generates switch control instructions based on the decision results and directly drives the actuators of the power distribution equipment through the high-speed I / O interface. The strategy table is generated through offline simulation training and contains the optimal operation sequence under different fault scenarios.

[0012] According to the present invention, a smart grid-based integrated anti-misoperation interlocking system for power distribution networks is provided. This system constructs a heterogeneous communication network by integrating the physical layers of 5G and optical fiber, and employs a dual-channel redundancy backup strategy to achieve dynamic switching of communication paths. Specifically, it includes the following steps: At key nodes of the power distribution network, 5G wireless communication modules and fiber optic wired interfaces are deployed simultaneously to form a 5G+fiber dual-channel transmission capability. The 5G channel adopts the 3GPP standard protocol, while the fiber optic channel adopts the IEC 61850 or DL / T 860 standard protocol. Data format unification is achieved through a protocol conversion gateway. Based on the geographical features and equipment distribution of the power distribution area, plan the coverage of 5G base stations and the layout of the fiber optic ring network to ensure that each terminal node can access at least one reliable communication method. According to the present invention, a smart grid-based integrated anti-misoperation interlocking system for power distribution networks includes a dual-channel redundancy backup strategy: Set the fiber channel as the default primary channel and the 5G channel as the backup channel. The communication terminal has a built-in channel quality assessment module that continuously monitors parameters such as signal strength, bit error rate, and transmission delay of the main channel. When a parameter of the main channel exceeds a preset threshold, the channel switching condition is triggered. Calculate the overall health score for both the main channel and the backup channel: H = w 1 S + w 2 (1 B )+ w 3 (1 L ) in, S This is the normalized value of the signal strength. B For bit error rate, L For delayed normalized values, w 1, w 2, w 3 represents the weighting coefficient; When the main channel scores H 主 Lower than the score of the backup channel H 备 And the difference exceeds the threshold Δ H At that time, initiate path migration.

[0013] According to the present invention, a smart grid-based integrated anti-misoperation interlocking system for distribution networks completes local logic judgment at edge computing nodes through FPGA programmable logic devices to shorten the processing latency of routine operation instructions, as shown below:

[0014] in, T edge For edge computing latency, T central For the latency of traditional centralized architecture, L data For data length, R central To centralize processing speed, B FPGAFor FPGA bandwidth, P parallel This is the parallel processing factor.

[0015] According to the present invention, a smart grid-based integrated anti-misoperation interlocking system for distribution networks, the dynamic interlocking rule base constructed from a knowledge graph satisfies logical constraints, as shown below:

[0016] Where e is a power grid topology edge, R topo R is a set of topological relation rules. safety For a set of security rules.

[0017] The topological reasoning of knowledge graphs satisfies a graph theory model, represented as:

[0018] Where G is the power grid topology graph, V is the node set, E is the edge set, and R is the relation set, containing topological relations. r topo and locking relationship r lock .

[0019] According to the present invention, a smart grid-based integrated anti-misoperation interlocking system for distribution networks is established by using digital twin technology to create a virtual image system of the power grid. The feasibility of the operation scheme is verified by Monte Carlo simulation, and the following formula is satisfied:

[0020] in, P pass To improve the success rate of the optimized operation plan, P original Given the original pass rate, the number of Monte Carlo simulations N satisfies:

[0021] Where Z is the quantile corresponding to the confidence level, and p is the estimated pass rate. This refers to the error range; The bidirectional synchronization of a digital twin system satisfies the following formula:

[0022] in, S physical For physical entity state, S digital For the state of the digital model, Δ t sync To synchronize delay, t threshold To allow the maximum delay.

[0023] According to the present invention, a smart grid-based integrated anti-misoperation interlocking system for distribution networks is constructed through a layered fault-tolerance mechanism to build a full life-cycle stability assurance system, including three levels: primary hardware-level circuit breaker protection, secondary resource-level dynamic auditing, and advanced expert-level problem-solving. This system satisfies the fault detection-isolation-self-healing closed-loop logic, as shown below:

[0024] in, D detect ( t Let be the fault detection function at time t. I isolate ( t + t 1) For t + t Time-1 fault isolation function, H heal ( t + t 2) For t + t The self-healing function at time 2, and t 1< t 2.

[0025] Therefore, compared with the existing technology, the intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks proposed in this invention has the following beneficial effects: 1. The intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks provided by this invention achieves decoupling of system functional modules through a hierarchical architecture of a distributed communication platform, an intelligent logic analysis engine, and a programmable execution control unit, thereby improving the flexibility and maintainability of the architecture.

[0026] 2. The intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks provided by this invention is based on a microservice architecture with modular deployment of functional components. It combines asynchronous message queues and domain event-driven mechanisms to reduce dependencies between components and enhance system scalability and stability.

[0027] 3. The intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks provided by this invention deploys edge computing nodes at key nodes such as substations and ring main units, and uses FPGA programmable logic devices to complete local logic judgments, reducing the processing latency of conventional operation instructions to 1 / 5 of that of the traditional centralized architecture, thus significantly improving real-time control efficiency.

[0028] 4. The intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks provided by this invention enables the edge computing cluster to respond quickly based on local information and simultaneously synchronize important data to the central control system, achieving an efficient mode of "local rapid processing + global collaborative decision-making".

[0029] 5. The intelligent grid-based integrated anti-misoperation interlocking system for distribution networks provided by this invention forms an adaptive decision-making model based on reinforcement learning, and combines natural language processing technology to parse the semantics of instructions and generate standardized operation sequences, thereby improving the level of intelligent decision-making.

[0030] 6. The intelligent grid-based integrated anti-misoperation interlocking system for distribution networks provided by this invention adopts a three-level fault-tolerant system of "intelligent circuit breaker mechanism (circuit breaker mode) + service health monitoring + abnormal flow control" to form a complete closed loop of "fault detection-isolation-self-healing", thereby improving the system's anti-interference capability and recovery efficiency.

[0031] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of an embodiment of the intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks according to the present invention.

[0033] Figure 2 This is a flowchart illustrating the microservice architecture, asynchronous message queue, and domain event-driven mechanism in an embodiment of the intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks according to the present invention.

[0034] Figure 3 This is a schematic diagram illustrating the process of completing local logic judgment using an FPGA programmable logic device in an embodiment of a smart grid-based integrated anti-misoperation interlocking system for power distribution networks according to the present invention.

[0035] Figure 4 This is a flowchart illustrating the instantaneous identification and decision-making process for emergency faults in an embodiment of an intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks according to the present invention.

[0036] Figure 5 This is a flowchart illustrating the heterogeneous communication network and dual-channel redundant backup strategy in an embodiment of the intelligent grid-based integrated anti-misoperation interlocking system for power distribution networks according to the present invention. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0038] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0039] An embodiment of a smart grid-based integrated anti-misoperation interlocking system for power distribution networks See Figure 1 to Figure 5 This embodiment provides a smart grid-based integrated anti-misoperation interlocking system for power distribution networks, including: The optimization module is used to implement a layered decoupling design of the system through a distributed communication platform, an intelligent logic analysis engine, and a programmable execution control unit. The layered decoupling relationship of the system is represented as follows:

[0040] in, d decouple <1, C sys For the overall coupling of the system, C modulei Let i be the independent coupling degree of the i-th module. d decouple These are the layered decoupling coefficients; The deployment module is used to modularly deploy functional components based on a microservice architecture. It utilizes an asynchronous message queue as a cross-component communication bus, combines a domain event-driven mechanism to reduce system coupling, and completes local logic judgments on edge computing nodes through FPGA programmable logic devices. The network hardening module is used to build a 5G+ fiber optic heterogeneous communication network and ensures communication quality through dual-channel redundant transmission technology. The intelligent decision-making module is used to form an adaptive decision-making model based on reinforcement learning and knowledge graph technology, and output a standardized operation sequence. O ={ o 1, o 2,…, o m}, represented as:

[0041] in, r t Let be the reward value at time t. c As a discount factor, V ( s t+1 ) is the state s t+1 The value function; The digital twin verification module is used to establish a virtual image system of the power grid through digital twin technology and to verify the feasibility of the operation scheme using Monte Carlo simulation.

[0042] In this embodiment, as Figure 2 As shown, a modular design of functional components is implemented based on a microservice architecture, and cross-component information interaction is completed through asynchronous message queues and domain event-driven mechanisms. Specifically, the steps include: Functional module division steps: The power system control function is divided into multiple independent small microservice units. Each microservice unit is responsible for specific functions or business logic, including but not limited to load monitoring, power balancing, equipment status monitoring, and fault early warning. Each microservice unit achieves interoperability through a lightweight interface REST API to avoid direct dependency calls and reduce system coupling.

[0043] Asynchronous message queue communication steps: An asynchronous message queue is used as a cross-component communication bus. Each microservice unit publishes or subscribes to events through the message queue to achieve loosely coupled communication. The message queue supports the transmission of multiple types of events, including status changes, control commands, and early warning information, to ensure the reliability and real-time performance of data transmission.

[0044] Domain event-driven mechanism: Define the domain event model, including event type, event source, event data, and triggering conditions; when a microservice unit detects a specific state change (such as device malfunction or overload), it generates the corresponding domain event and publishes it to the message queue.

[0045] In this embodiment, as Figure 3 As shown, local logic decisions are performed at the edge computing node using FPGA programmable logic devices, specifically including the following steps: Integrate FPGA programmable logic devices in edge computing nodes and configure them as hardware acceleration units independent of the main processor; The FPGA internally implements logic circuit partitioning, including a data acquisition module, a fault detection module, a control command generation module, and a communication interface module; each module is interconnected through the FPGA's internal high-speed bus, supporting nanosecond-level data transmission and synchronization.

[0046] The FPGA is directly connected to the power distribution feeder sensors (such as current transformers and voltage transformers) through a high-speed ADC interface to collect three-phase current and voltage waveform data in real time. Data preprocessing logic is implemented inside the FPGA, including filtering, denoising, feature extraction (such as zero-sequence current component and harmonic content), and generating standardized detection parameters.

[0047] In this embodiment, as Figure 4 As shown, it also performs instantaneous identification and decision-making for emergency faults: Single-phase grounding fault detection: Parallel comparison circuits are deployed inside the FPGA to monitor the amplitude and phase of the zero-sequence current in real time; when the zero-sequence current exceeds a preset threshold (e.g., 5A for 10ms), the fault flag is triggered, and the time of the fault occurrence and the feeder topology location are recorded at the same time. Short circuit fault detection: The FPGA calculates the current mutation rate (di / dt) using a differential algorithm. When the mutation rate exceeds a threshold (e.g., 10kA / ms), it is immediately identified as a short circuit fault. The FPGA calls the corresponding action (such as cutting off the faulty branch or isolating adjacent devices) from the pre-stored strategy table according to the fault type and feeder topology. The strategy table is generated through offline simulation training and contains the optimal operation sequence under different fault scenarios.

[0048] The FPGA generates switching control commands (such as circuit breaker tripping signals) based on the decision results, and directly drives the actuators of the power distribution equipment through a high-speed I / O interface. The command generation and execution process is completed entirely within the FPGA, without waiting for authorization from the remote master station. At the same time, the FPGA sends fault reports (including timestamps, fault types, and operation results) to the master station through an asynchronous message queue for subsequent analysis and recording.

[0049] In a ring main unit site, the FPGA synchronously monitors the status parameters of multiple power distribution feeders and constructs a local topology model. When a feeder fails, the FPGA quickly identifies the faulty branch and automatically calculates the minimum isolation range (such as disconnecting the faulty branch while maintaining power supply to other feeders). Through the FPGA's internal parallel processing capabilities, it enables synchronous operation of multiple branches, preventing the fault from spreading to the entire ring network.

[0050] As can be seen, compared with the latency of the traditional architecture: under the traditional architecture, fault data needs to be uploaded to the remote master station (latency 10-50ms), and the master station issues control commands (latency 10-30ms), with a total response time of 20-80ms; the present invention reduces the response time to <1ms (including fault detection and command execution) through local processing on FPGA, and improves efficiency by 10-100 times.

[0051] In this embodiment, as Figure 5 As shown, a heterogeneous communication network is constructed by integrating the physical layers of 5G and optical fiber, and a dual-channel redundancy backup strategy is adopted to achieve dynamic switching of communication paths. Specifically, the following steps are included: At key nodes of the power distribution network (such as substations, ring main units, and power distribution terminals), 5G wireless communication modules and fiber optic wired interfaces are deployed simultaneously to form a 5G+fiber dual-channel transmission capability. The 5G channel adopts 3GPP standard protocols (such as NB-IoT or 5GNR), while the fiber optic channel adopts IEC 61850 or DL / T 860 standard protocols. Data format unification is achieved through a protocol conversion gateway. Based on the geographical features and equipment distribution of the power distribution area, plan the coverage of 5G base stations and the layout of the fiber optic ring network to ensure that each terminal node can access at least one reliable communication method.

[0052] In this embodiment, the dual-channel redundancy backup strategy includes: Primary and backup channel definition: Set the fiber optic channel as the default primary channel and the 5G channel as the backup channel (or vice versa, dynamically configured according to the scenario). The communication terminal has a built-in channel quality assessment module to continuously monitor parameters such as signal strength (RSSI), bit error rate (BER), and transmission delay (Latency) of the main channel; When a parameter of the main channel is detected to exceed a preset threshold (such as RSSI < -90dBm, BER > 10), -5 When the delay is greater than 50ms, the channel switching condition is triggered.

[0053] Calculate the overall health score for both primary and backup channels. H = w 1 S + w 2 (1 B )+ w 3 (1 L ) in, S This is the normalized value of the signal strength. B For bit error rate, L For delayed normalized values, w 1, w 2, w 3 represents the weighting coefficient; When the main channel scores H 主 Lower than the score of the backup channel H 备 And the difference exceeds the threshold Δ H When switching occurs, path migration is initiated. The communication terminal maintains primary and backup channel connections simultaneously through dual protocol stacks; during switching, only the data routing table needs to be modified, without re-establishing the session.

[0054] In this embodiment, local logic judgments are performed at the edge computing node using FPGA programmable logic devices to shorten the processing latency of conventional operation instructions, as shown below:

[0055] in, T edge For edge computing latency, Tcentral For the latency of traditional centralized architecture, L data For data length, R central To centralize processing speed, B FPGA For FPGA bandwidth, P parallel This is the parallel processing factor.

[0056] In this embodiment, a 5G+fiber heterogeneous communication network is constructed, and communication quality is ensured through dual-channel redundant transmission technology. The dynamic link selection algorithm satisfies the following: R reliability ≥99% and |Δ T jitter |≤5ms; in, R reliability For communication reliability, Δ T jitter The delay jitter value is given, and the channel switching condition is as follows:

[0057] in, SNR 5G For 5G channel signal-to-noise ratio. c threshold This is a preset threshold.

[0058] In this embodiment, the dynamic locking rule base constructed from the knowledge graph satisfies logical constraints, as shown below:

[0059] Where e is a power grid topology edge, R topo R is a set of topological relation rules. safety For a set of security rules.

[0060] The topological reasoning of knowledge graphs satisfies a graph theory model, represented as:

[0061] Where G is the power grid topology graph, V is the node set, E is the edge set, and R is the relation set (containing topological relations). r topo and locking relationship r lock ).

[0062] In this embodiment, a virtual image system of the power grid is established using digital twin technology, and the feasibility of the operation scheme is verified using Monte Carlo simulation, satisfying the following formula:

[0063] in,P pass To improve the success rate of the optimized operation plan, P original Given the original pass rate, the number of Monte Carlo simulations N satisfies:

[0064] Where Z is the quantile corresponding to the confidence level, and p is the estimated pass rate. This represents the error range.

[0065] The bidirectional synchronization of a digital twin system satisfies the following formula:

[0066] in, S physical For physical entity state, S digital For the state of the digital model, Δ t sync To synchronize delay, t threshold To allow the maximum delay.

[0067] In this embodiment, the semantic parsing of the adaptive decision model satisfies:

[0068] in, I command For input instructions, O sequence For the generated operation sequence, NLP ( ) is a natural language processing function.

[0069] In this embodiment, the hierarchical decision-making mechanism satisfies the data synchronization rules:

[0070] in, D edge For edge node data, D central For data from the central control system, D critical This is a subset of the key data.

[0071] Furthermore, digital twin systems provide two-way real-time synchronization between physical entities and digital models, ensuring that the virtual environment accurately reflects the actual operating status of the power grid.

[0072] Furthermore, the operation simulation sandbox environment supports simulation and deduction under various complex scenarios, which can significantly improve the pass rate of operation schemes to more than 40% on the original basis.

[0073] Furthermore, edge computing nodes are deployed in key network nodes such as substations and ring main units, reducing the processing latency of routine operation instructions to 1 / 5 of that of traditional centralized architectures.

[0074] Furthermore, the dynamic link selection algorithm for heterogeneous communication networks using 5G+fiber can automatically switch communication channels based on real-time channel quality assessment results, ensuring a communication reliability of no less than 99% and controlling latency jitter within ±5ms.

[0075] In this embodiment, a full lifecycle stability assurance system is constructed through a layered fault tolerance mechanism, including three levels: primary hardware-level circuit breaker protection, secondary resource-level dynamic auditing, and advanced expert-level problem-solving. The three-level fault tolerance system satisfies the closed-loop logic of fault detection-isolation-self-healing.

[0076] in, D detect ( t Let be the fault detection function at time t. I isolate ( t + t 1) For t + t Time-1 fault isolation function, H heal ( t + t 2) For t + t The self-healing function at time 2, and t 1< t 2.

[0077] Specifically, the fault-tolerant mechanism, designed through a layered approach, constructs a primary hardware-level circuit breaker protection system, a secondary resource-level dynamic audit system, and a high-level expert-level system for handling complex issues. These layers work together to mitigate risks, prevent bottlenecks, and handle complex faults. The specific implementation steps are as follows: Primary protection: Hardware-level automatic fuse isolation. In the communication links and control loops of critical power equipment (such as circuit breakers, ring main unit switches, and distributed power supply interfaces), electronic fuses with self-detection functions are integrated. The fuse parameters are configured according to the importance of the equipment. For example, the fusing threshold for primary equipment (main power circuit) is set to 150% current overload for 200ms, and for secondary equipment (backup branch) it is set to 200% current overload for 100ms. When the fuse detects an anomaly such as a short circuit, overvoltage, or communication interruption, it immediately disconnects the physical / logical connection between the faulty equipment and the system to prevent the fault from spreading to adjacent equipment. At the same time, it triggers a local alarm through a hardware interrupt signal and encapsulates the fault type, timestamp, and device ID into a standard event and reports it to the secondary audit layer.

[0078] Secondary protection: Resource-level dynamic auditing and bottleneck prevention: Deploy resource auditing modules on edge computing nodes to scan system resource utilization (CPU, memory, bandwidth), device status (temperature, load rate), and communication quality (latency, packet loss rate) at fixed intervals (e.g., every 5 minutes). Audit data is used to predict resource consumption trends through time series analysis algorithms (e.g., ARIMA model) to identify potential bottlenecks (e.g., a feeder load rate consistently above 85%). When auditing detects that resource utilization exceeds a safety threshold (e.g., CPU > 90% for 3 consecutive cycles), a load balancing mechanism is automatically triggered: non-critical tasks (e.g., historical data storage) are migrated to low-load nodes; the priority of critical tasks (e.g., fault detection) is adjusted and additional computing resources are allocated; simultaneously, a preventative maintenance work order is generated and pushed to the operations and maintenance terminal to guide on-site repairs.

[0079] Advanced protection: Expert-level decision support for complex issues: Establish a remote expert decision-making center, integrating multi-domain knowledge bases (including power equipment manuals, historical fault cases, and simulation models) and AI-assisted analysis tools (such as Bayesian networks and deep learning fault classifiers); and support remote expert consultation by accessing field data (waveform graphs, equipment logs, and audit reports) in real time via 5G / fiber optic dual channels.

[0080] The system will automatically upgrade to the expert diagnostic room when it encounters the following scenarios: Complex faults that cannot be resolved by primary circuit breaking and secondary auditing (such as communication interruption + equipment overload). Unknown anomalies caused by new equipment or newly connected loads; Multiple nodes report errors simultaneously without a clear causal relationship; Experts retrieve data and run simulations through an interactive interface, and finally output decision instructions (such as temporarily adjusting protection settings or isolating specific areas), while simultaneously updating the knowledge base.

[0081] In practical applications, if the fault is not resolved within 10 seconds after the primary fuse is triggered, it is automatically escalated to the secondary audit layer. If the secondary audit layer times out (e.g., not resolved within 30 minutes) or is identified as a complex fault, it is escalated to the expert diagnostic room. After the experts make a decision, they provide feedback on the correction strategy to the primary / secondary layers, forming a closed-loop optimization.

[0082] In summary, the system provided in this embodiment constructs a layered and decoupled design through key technical components such as a distributed communication platform, an intelligent logic analysis engine, and edge computing nodes, thereby reducing the overall coupling of the system and shortening the instruction processing latency.

[0083] First, functional components are deployed based on a microservice architecture to achieve modular design. At the same time, asynchronous message queues are used as cross-component communication buses combined with domain event-driven mechanisms to complete information interaction. Specifically, in this stage, the power system control functions are divided into multiple independent small microservice units. Each service is responsible for specific functions or business logic, such as load monitoring or power balancing. These microservices interoperate through lightweight interfaces (REST APIs) during operation, and domain events notify each service of relevant state changes, avoiding direct dependency calls that would increase complexity. For example, when a power outage warning is triggered by an abnormality in a substation device, the warning module will issue a domain event signal to notify all relevant upstream and downstream microservices to immediately start the response process without having to call them one by one. This reduces the hard coupling between modules.

[0084] Secondly, using FPGA programmable logic devices at edge computing nodes to process local logic reduces response time. Specifically, this approach enables urgent operations such as short-circuit detection to be executed quickly near the data source instead of having to return to a remote data center to receive instructions. This transforms the previous requirement of waiting for tens of milliseconds of feedback into near real-time decision-making and action. For example, in a ring main unit site with multiple power distribution feeders, once a single-phase ground fault occurs and reaches a preset threshold condition, the FPGA can instantly identify and quickly disconnect the power supply to the faulty branch without having to request permission from the remote master station before operating. This significantly reduces the possibility of potential hazard spread and saves several times the processing time of traditional architectures.

[0085] To further enhance communication stability and ensure reliability targets are met, a highly available heterogeneous communication network combining 5G and fiber optics is introduced at the physical transmission layer. A dual-channel redundancy backup strategy is also added. Specifically, if signal attenuation or increased latency on a certain line exceeds specified limits, the system dynamically assesses the current channel health based on a pre-configured algorithm and promptly migrates to another high-performing path. Conversely, it continues operating stably as is. For example, if in some remote areas only a single low-speed wireless band connects to the core gateway server, but heavy rain causes frequent packet loss due to radio interference, the system can seamlessly switch to a backup high-bandwidth fiber optic line to re-establish an efficient and reliable connection and restore normal communication. This ensures continuous, fault-free operation within the system's defined critical service level objectives of availability (up to three 9s) and precise synchronization within ±5ms.

[0086] Then, the adaptive decision model is designed to generate an operation sequence scheme that meets the optimal choice under the current environment. With the help of reinforcement learning, the internal parameters are continuously adjusted to better fit the needs of the actual scenario. In addition, after integrating natural language processing devices, it can accurately extract effective keyword elements from the task briefs issued by humans and translate them into a command form that is easy for computers to understand and express for subsequent execution. For example, when facing complex power grid reconfiguration tasks, such as changing the position relationship of transformer contacts to alleviate the overload pressure during peak electricity consumption, it can analyze historical experience databases and real-time load curve charts to comprehensively consider the influence weight of multiple factors and finally select the most reasonable switch combination arrangement mode, which is then handed over to the operators for verification and confirmation before formal implementation of the modification.

[0087] At the same time, a complete three-level fault-tolerant framework has been built to ensure that the stability and risk resistance of the network remain at their best throughout its entire lifecycle: Primary protection measures include equipping certain critical components with automatic fuses to prevent potential cascading effects. The secondary level adds a comprehensive tracking and scanning process to monitor the overall resource allocation, and periodic audit procedures identify potential bottlenecks and prevent and correct them in advance. The final link involves an expert diagnostic lab that specializes in resolving exceptional and complex cases that defy conventional judgment. This establishes a comprehensive closed-loop solution encompassing fault prediction, early warning, handling, and recovery. Specifically, when an area is suddenly threatened by a strong external magnetic attack that causes electronic components to fail, an emergency cutoff circuit is activated to prevent damage to other related circuits and ensure their continued normal operation. Simultaneously, a backup power supply is activated to temporarily replace damaged components, maintaining critical functions for a period until professional personnel arrive to thoroughly repair and replace the hardware, fully restoring all original functions.

[0088] Finally, by using digital twin technology to create a comprehensive and detailed digital version that maps the layout and features of the corresponding physical equipment, it is possible to conduct various hypothetical calculations, drills, and training processes safely, efficiently, and reliably achieve the goals, resulting in a significant improvement in the achievement rate. For example, before scheduling daily planned maintenance shifts, this simulation test space can be used to simulate the linkage effects between units under various complex operating conditions, identify potential hazards, and rationally allocate human and material resources to save costs. According to calculations, this method improves efficiency by at least 40% compared to previous arrangements based solely on experience and subjective intuition.

[0089] By designing a distributed communication platform, an intelligent logic analysis engine, and a programmable execution control unit, the system achieves layered decoupling. Functional components are deployed modularly according to a microservice architecture, and an asynchronous message queue is used as a cross-component communication bus. The coupling of the system is significantly reduced by leveraging a domain event-driven mechanism. FPGA programmable logic devices are introduced in edge computing nodes to complete local logic judgments. This solution enables routine operation instructions to be processed quickly near the data source, effectively reducing latency to 1 / 5 of that of traditional centralized architectures.

[0090] Meanwhile, to ensure communication reliability, the architecture proposes to construct a 5G+fiber heterogeneous communication network and adopt dual-channel redundant transmission technology. Combined with a dynamic link selection algorithm, it automatically switches the optimal communication channel based on the channel quality assessment results, ensuring that the overall power grid communication reliability reaches no less than 99.9%, and accurately controlling the latency jitter within ±5ms. On this basis, the architecture introduces an adaptive decision-making model formed by reinforcement learning and knowledge graph technology. While standardizing the output operation sequence, it provides a safety assessment report and uses natural language processing technology to analyze the semantic features of operation instructions to improve the accuracy of decision-making. At the same time, the dynamic generation function of knowledge graph reasoning of power grid topology relationships and locking rules further enhances the system's flexibility and security.

[0091] Furthermore, digital twin technology, as a crucial component of the entire architecture, not only constructs a virtual image system of the power grid, supporting multi-scenario simulation and strategy verification, but also employs methods such as Monte Carlo simulation verification to ensure the feasibility and reliability of pre-operation plans. These combined steps significantly improve the overall system's intelligent response level and actual operation pass rate, exceeding traditional solutions by at least 40%. The hierarchical decision-making mechanism enables the edge computing cluster to possess autonomous and rapid response capabilities while also feeding necessary information back to the central control system for global optimization management. Overall, this optimized architecture effectively achieves multiple technical goals: reducing system coupling, improving command response efficiency, enhancing communication reliability, and ensuring high feasibility.

[0092] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0093] The above embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. Any non-substantial changes and substitutions made by those skilled in the art based on the present invention shall fall within the scope of protection claimed by the present invention.

Claims

1. A power distribution network intelligent meshing comprehensive anti-error locking system, characterized in that, include: The optimization module is used to implement a layered decoupling design of the system through a distributed communication platform, an intelligent logic analysis engine, and a programmable execution control unit. The layered decoupling relationship of the system is represented as follows: in, δ decouple <1, C sys For the overall coupling of the system, C modulei Let i be the independent coupling degree of the i-th module. δ decouple These are the layered decoupling coefficients; The deployment module is used to modularly deploy functional components based on a microservice architecture. It utilizes an asynchronous message queue as a cross-component communication bus, combines a domain event-driven mechanism to reduce system coupling, and completes local logic judgments on edge computing nodes through FPGA programmable logic devices. The network hardening module is used to build a heterogeneous communication network of 5G and fiber optics, and ensures communication quality through dual-channel redundant transmission technology. The intelligent decision-making module is used to form an adaptive decision-making model based on reinforcement learning and knowledge graph technology, and output a standardized operation sequence. O ={ o 1, o 2,…, o m }, represented as: in, r t Let be the reward value at time t. γ As a discount factor, V ( s t+1 ) is the state s t+1 The value function; The digital twin verification module is used to establish a virtual image system of the power grid through digital twin technology and to verify the feasibility of the operation scheme using Monte Carlo simulation.

2. The system according to claim 1, characterized in that, The modular design of functional components is implemented based on a microservice architecture, and cross-component information interaction is accomplished through asynchronous message queues and domain event-driven mechanisms. The specific steps include: Functional module division steps: The power system control function is divided into multiple independent small microservice units. Each microservice unit is responsible for a specific function or business logic, including at least load monitoring, power balancing, equipment status monitoring, and fault early warning. Each microservice unit achieves interoperability through a lightweight interface REST API. Asynchronous message queue communication steps: An asynchronous message queue is used as a cross-component communication bus. Each microservice unit publishes or subscribes to events through the message queue to achieve loosely coupled communication. The message queue supports the transmission of multiple types of events, including at least status changes, control commands, and warning information. Domain event-driven mechanism: Define a domain event model, including at least event type, event source, event data, and triggering conditions; when a microservice unit detects a specific state change, it generates a corresponding domain event and publishes it to the message queue.

3. The system according to claim 1, characterized in that, Local logic decisions are performed at edge computing nodes using FPGA programmable logic devices, specifically including the following steps: Integrate FPGA programmable logic devices in edge computing nodes and configure them as hardware acceleration units independent of the main processor; The FPGA internally implements logic circuit partitioning, including a data acquisition module, a fault detection module, a control command generation module, and a communication interface module; The FPGA is directly connected to the power distribution feeder sensor via a high-speed ADC interface to acquire three-phase current and voltage waveform data in real time. Implement data preprocessing logic within the FPGA, including at least filtering, noise reduction, and feature extraction, to generate standardized detection parameters.

4. The system according to claim 3, characterized in that, It also performs instantaneous identification and decision-making for emergency failures: Single-phase grounding fault detection: Parallel comparison circuits are deployed inside the FPGA to monitor the amplitude and phase of the zero-sequence current in real time; when the zero-sequence current exceeds the preset threshold, the fault flag is triggered, and the time of the fault occurrence and the feeder topology location are recorded at the same time. Short circuit fault detection: The FPGA calculates the current mutation rate through a differential algorithm. When the mutation rate exceeds the threshold, it is immediately determined to be a short circuit fault. Based on the fault type and feeder topology, the FPGA calls the corresponding action from the pre-stored strategy table. The FPGA generates switch control instructions based on the decision results and directly drives the actuators of the power distribution equipment through the high-speed I / O interface. The strategy table is generated through offline simulation training and contains the optimal operation sequence under different fault scenarios.

5. The system according to claim 1, characterized in that, A heterogeneous communication network is constructed by integrating the physical layers of 5G and optical fiber, and a dual-channel redundancy backup strategy is adopted to achieve dynamic switching of communication paths. The specific steps include: At key nodes of the power distribution network, 5G wireless communication modules and fiber optic wired interfaces are deployed simultaneously to form a 5G+fiber dual-channel transmission capability. The 5G channel adopts the 3GPP standard protocol, while the fiber optic channel adopts the IEC 61850 or DL / T 860 standard protocol. Data format unification is achieved through a protocol conversion gateway. Based on the geographical features and equipment distribution of the power distribution area, plan the coverage of 5G base stations and the layout of the fiber optic ring network to ensure that each terminal node can access at least one reliable communication method.

6. The system according to claim 5, characterized in that, Dual-channel redundancy backup strategies include: Set the fiber channel as the default primary channel and the 5G channel as the backup channel. The communication terminal has a built-in channel quality assessment module that continuously monitors parameters such as signal strength, bit error rate, and transmission delay of the main channel. When a parameter of the main channel exceeds a preset threshold, the channel switching condition is triggered. Calculate the overall health score for both the main channel and the backup channel: H = w 1 S + w 2 (1 B )+ w 3 (1 L ) in, S This is the normalized value of the signal strength. B For bit error rate, L For delayed normalized values, w 1, w 2, w 3 represents the weighting coefficient; When the main channel scores H 主 Lower than the score of the backup channel H 备 And the difference exceeds the threshold Δ H At that time, initiate path migration.

7. The system according to any one of claims 1 to 6, characterized in that: At edge computing nodes, local logic decisions are performed using FPGA programmable logic devices to shorten the processing latency of routine operation instructions, as shown below: in, T edge For edge computing latency, T central For the latency of traditional centralized architecture, L data For data length, R central To centralize processing speed, B FPGA For FPGA bandwidth, P parallel This is the parallel processing factor.

8. The system according to any one of claims 1 to 6, characterized in that: The dynamic locking rule base constructed from the knowledge graph satisfies logical constraints, as shown below: wherein e is an electrical grid topology edge, R topo is a set of topology relationship rules, R safety is a set of security rules; The topological reasoning of knowledge graphs satisfies a graph theory model, represented as: Where G is the power grid topology graph, V is the node set, E is the edge set, and R is the relation set, containing topological relations. r topo and locking relationship r lock .

9. The system according to any one of claims 1 to 6, characterized in that: A virtual image system of the power grid is established using digital twin technology, and the feasibility of the operation scheme is verified using Monte Carlo simulation, satisfying the following formula: in, P pass To improve the success rate of the optimized operation plan, P original Given the original pass rate, the number of Monte Carlo simulations N satisfies: Where Z is the quantile corresponding to the confidence level, and p is the estimated pass rate. This refers to the error range; The bidirectional synchronization of a digital twin system satisfies the following formula: in, S physical For physical entity state, S digital For the state of the digital model, Δ t sync To synchronize delay, t threshold To allow the maximum delay.

10. The system according to any one of claims 1 to 6, characterized in that: A layered fault-tolerance mechanism is used to construct a full lifecycle stability assurance system, including three levels: primary hardware-level circuit breaker protection, secondary resource-level dynamic auditing, and advanced expert-level complex problem-solving. This system satisfies the closed-loop logic of fault detection-isolation-self-healing, as shown below: in, D detect ( t Let be the fault detection function at time t. I isolate ( t + τ 1) For t + τ Time-1 fault isolation function, H heal ( t + τ 2) For t + τ The self-healing function at time 2, and τ 1< τ 2.

Images