Intelligent disassembling and direct remote control method for distribution network dispatching comprehensive instruction

Abstract

The invention relates to the technical field of distribution network scheduling, and discloses a distribution network scheduling comprehensive instruction intelligent disassembly and direct remote control method, which comprises the following steps of: constructing a digital distribution network model, a dynamic node branch type topological structure, a dual rule library and an equipment comprehensive state library; the method comprises the following steps: analyzing an instruction core demand through entity identification, and performing three-level disassembly and verification to form a candidate instruction sequence; performing local state initial judgment, whole network topology analysis and whole process verification; quantifying and grading risks, and screening qualified instructions in combination with permission verification; and issuing an instruction according to a topological sequence, confirming completeness through double check, synchronously updating data and carrying out secondary false alarm prevention. According to the method, a standard operation sequence is efficiently generated through BERT three-level disassembly instructions in combination with multi-level anti-misoperation verification and linkage risk and authority control by depending on topology dynamic updating and direct remote control, the misoperation risk is reduced, and the distribution network scheduling safety and efficiency are balanced.

Description

Technical Field

[0001] This invention relates to the field of distribution network dispatching technology, and more specifically, to a method for intelligent decomposition and direct remote control of integrated distribution network dispatching instructions. Background Technology

[0002] With the continuous expansion of the distribution network, the large-scale integration of distributed power sources, and the diversified growth of electricity load, the distribution network structure is becoming increasingly complex, leading to an explosive growth in dispatching workload. As the brain of the power system, distribution network dispatching must handle routine tasks such as issuing engineering orders and permits, as well as respond quickly to emergencies such as fault repairs, directly impacting power supply reliability and user experience. To alleviate dispatching pressure, the industry is gradually exploring optimization methods such as networked dispatching and segmented command structures.

[0003] However, existing technologies still fail to address the core pain points: traditional integrated commands rely on manual decomposition or are simply divided into blocks according to physical dimensions, making it impossible to understand the semantic connotations of ambiguous commands. This can easily lead to problems such as chaotic decomposition logic and omission of operation items, and it is difficult to adapt to the collaborative operation requirements of multiple devices under complex topologies. In direct remote control operation scenarios, because it is impossible to predict in advance the chain reaction of the operation on the entire network topology, it is difficult to effectively prevent serious violations such as closing the circuit with the ground wire or pulling the disconnect switch under load, and it cannot meet the high safety requirements of direct remote control. Summary of the Invention

[0004] To address the aforementioned technical issues, this invention provides a method for intelligent decomposition and direct remote control of integrated distribution network scheduling instructions. By using BERT three-level instruction decomposition, combined with multi-layered error prevention verification, and linking risk and access control, and relying on dynamic topology updates and direct remote control, it efficiently generates standard operation sequences, reduces the risk of misoperation, and balances the security and efficiency of distribution network scheduling.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows:

[0006] A method for intelligent decomposition and direct remote control of integrated instructions for distribution network dispatch includes the following steps: S1. Mapping the primary wiring structure of physical distribution network lines to a digital distribution network line model using a standardized object-oriented modeling tool; calling a topology analysis algorithm based on the digital model to generate a node-branch type topology structure containing topology nodes, topology branches, and topology adjacency matrices, which is automatically updated synchronously when wiring changes; building a dual rule base containing a remote control operation safety constraint rule base and a mis-reasoning rule base, embedding an intelligent mis-reasoning algorithm; deploying entity recognition technology to integrate the status variables of primary and secondary equipment and offline identification information, and identifying the basic and special states of equipment through preset logic to form a dynamically updated comprehensive equipment status base;

[0007] S2. Analyze the core requirements of the scheduling integrated command through entity recognition technology. Based on the node branch topology, the basic operation command is disassembled into three levels: region, bus, and equipment. The equipment integrated status library is called to verify the matching between the disassembly command and the equipment status, eliminate conflicting commands, and form a candidate command sequence.

[0008] S3. Taking the candidate instruction sequence as input, first delineate the area directly associated with the device to be operated, and complete the initial judgment of the device status in the area based on the comprehensive device status database; then conduct in-depth analysis of the entire network topology through a digital model, and integrate the results of the two-level analysis; and then conduct full-process verification of drafting, review, and pre-execution.

[0009] S4. Input the topology analysis results of S3 and the equipment status data of S1 into the real-time risk assessment algorithm to quantify the risk of instructions and manage them in a hierarchical manner. With the risk classification results as a prerequisite, call the management rule error prevention logic of S1, and verify the operator's permission level, duty period and authorized equipment set through the permission verification algorithm. Only instructions with risk levels matching permissions are allowed to enter the execution stage.

[0010] S5. Issue instructions that have passed the S4 verification in the topology association order, and verify the completion of the operation through automatic telemetry and teleindication or video-assisted manual verification. After each operation, update the real-time data of the entire network and the node branch topology of S1 synchronously, and re-input the updated data into the intelligent anti-misoperation reasoning algorithm of S1 and the risk assessment algorithm of S4. If an anomaly is determined, block subsequent instructions and issue an alarm.

[0011] As a preferred embodiment of the present invention, the specific steps for constructing the digital distribution network line model in S1 are as follows:

[0012] Using a Public Information Modeling (CIM) standardized object-oriented modeling tool conforming to the IEC 61970 standard, physical distribution network primary equipment is transformed into structured digital instances through equipment attribute mapping functions. The expression of the mapping function is as follows: ;

[0013] in, For the first A complete set of attributes for each physical device, including device type, physical connection port location information, rated electrical parameters, and globally unique device number; The device is assigned a globally unique digital identifier, using a 16-bit string encoding rule. The first 8 bits are the administrative division code of the region to which the device belongs, and the last 8 bits are the device's sequential serial number. Equipment type is coded according to the following rules: 01=busbar, 02=circuit breaker, 03=disconnector, 04=load, 05=generator, 06=transformer, 07=grounding switch. The codes are unique and cannot be repeated. This is a set of three-dimensional coordinates of the electrical connection ports of the device, in the format of... The coordinate accuracy is 0.1m, which is used to accurately characterize the physical connection relationship between devices; This is a set of rated parameters for the equipment, including the rated voltage. Rated current Rated power ; The inherent impedance parameter of the equipment is calculated from the resistivity of the equipment material, its structural dimensions, and its temperature coefficient. The calculation formula is as follows: ,in The standard resistivity of the conductive material of the equipment, For the effective length of the conductive part, Cross-sectional area of ​​the conductive part For material temperature coefficient, This represents the difference between the actual ambient temperature and the standard temperature of 20℃.

[0014] Iterate through all primary devices in the distribution network, converting them one by one into digital device instances, and integrate them to form a digital distribution network line model: ;

[0015] in The total number of primary distribution network devices is represented by the digital distribution network line model, which is synchronized with the distribution automation system in real time with a data update cycle of no more than 1 second, thus realizing a one-to-one mapping between physical devices and digital models.

[0016] As a preferred embodiment of the present invention, the generation of the node-branch type topology structure in the construction of the basic model and support system described in S1 includes the following specific steps:

[0017] Topology node extraction: Core topology nodes are filtered using a topology node extraction function. The function expression is:

[0018] ;

[0019] in, Refers to core equipment such as busbars, circuit breakers, disconnect switches, generators, and transformers. Refers to loads and grounding switch equipment; generates a set of topology nodes. , This represents the total number of topology nodes. It serves as a globally unique identifier for the node, and is associated with the device. There is an associated mapping relationship;

[0020] Topology branch extraction: Standardized branches are generated through topology branch extraction and isometry functions. The function expression is as follows:

[0021]

[0022] in, Used as the unique identifier for a branch road; For branch equivalent impedance, single-port devices are calculated based on node-to-ground equivalent impedance. Two-port devices are calculated based on series or parallel impedance; series impedance is... , in parallel The three-port device is disassembled into three independent node-to-node branches and the equivalent impedance is calculated for each branch. Topology constraint clauses corresponding to branches, and remote operation safety constraint rule base. Match each rule in the table;

[0023] Construction of the topological adjacency matrix: The matrix expression is as follows:

[0024]

[0025] in, Represents a constrained, disconnected branch, with a value of To satisfy computational logic;

[0026] Dynamic topology update: Topology synchronization update is triggered by a wiring change identification function. The function expression is:

[0027]

[0028] in, The threshold for determining wiring changes is calculated based on the normalized change in equipment attributes; when When the time comes, the topology update process is automatically triggered, generating the updated topology structure. Ensure that the topology is consistent with the actual distribution network wiring.

[0029] As a preferred embodiment of the present invention, the construction of the dual rule base in the basic model and supporting system construction includes the following specific steps:

[0030] Construction of a remote control operation safety constraint rule base: Hierarchical control of constraint rules is achieved through a rule priority quantification function, the function expression of which is:

[0031]

[0032] in:

[0033] For the weighting coefficients, satisfying The default value is the risk level weight. Historical trigger frequency weight Scope of influence weight It can be dynamically adjusted according to the needs of power distribution network operation and maintenance;

[0034] For the first The safety constraint rules cover all scenarios of violations in distribution network operations, such as prohibiting the closing of grounding switches while the circuit is energized and prohibiting the pulling of disconnect switches under load.

[0035] The value ranges from 0 to 1, representing the operational risk level corresponding to the rule. Specifically, the value is 1.0 for the highest risk level, 0.9 for the first level, 0.7 for the second level, and 0.5 for the third level.

[0036] The weight for the historical trigger frequency of the rule, with a value range of 0-1, is calculated as follows: Regularly updated;

[0037] The weight of the impact range after the rule is triggered is 0-1, and is divided into levels according to the number of affected users: 1.0 for 100,000 users and above, 0.8 for 1.1 million users and 0.5 for less than 10,000 users.

[0038] Based on the above functions, constraint rules for all scenarios of distribution network operations are sorted out, forming a safety constraint rule library with priority identifiers. , To ensure the total number of security constraint rules, the rule base synchronously stores the topology node / branch association information corresponding to each rule;

[0039] Construction of the anti-misoperation reasoning rule base: It integrates five layers of anti-misoperation logic, including ontology state, relation state, device attributes, operation sequence, and safety distance, and embeds a fuzzy comprehensive evaluation-type intelligent anti-misoperation reasoning algorithm. The algorithm expression is as follows: ;

[0040] in: To prevent misjudgment of the result vector, For the operation belonging to the first The membership degree of the anti-misconduct category, with a value ranging from 0 to 1. To prevent accidental violations, the total number of violation categories;

[0041] The weight vector of error-prevention factors is determined by the analytic hierarchy process (AHP) and satisfies the following conditions: The default value is the original state. Relationship status Device attributes Operation sequence Safe distance ;

[0042] It is a fuzzy relation matrix. For the first The error prevention factor for the first The membership degree of the anti-misconduct category, with a value range of 0-1, is jointly determined by historical fault data and industry expert experience, and is regularly verified and updated.

[0043] For fuzzy synthesis operators, a weighted average operator is used, and the specific calculation formula is as follows: This ensures a balanced integration of multiple error-prevention factors.

[0044] As a preferred embodiment of the present invention, the construction of the comprehensive equipment status library in the construction of the basic model and support system described in S1 includes the following specific steps:

[0045] Multi-source state information fusion integrates state data from multiple heterogeneous devices through a state normalization function, the expression of which is: ;

[0046] in: To integrate the weighting coefficients, satisfy the following conditions: The default value is the weight of the primary device state quantity. Secondary equipment status quantity weights Offline signage information weight ;

[0047] For the device to be identified, and the digital distribution network line model The device instances in the database correspond one-to-one;

[0048] This is a primary equipment status variable, with a normalized value range of 0-1, defined as: closed state = 1, open state = 0, abnormal state = 0.5;

[0049] For secondary equipment status variables, the normalized value range is 0-1, defined as: protection device normal = 1, alarm status = 0.5, fault status = 0;

[0050] For offline signage information, the normalized value range is 0-1, defined as: Operation sign = 1, Maintenance sign = 0, Test sign = 0.3;

[0051] Device status determination includes basic status determination and special status determination. Basic status determination is implemented through a basic status mapping function, the expression of which is:

[0052]

[0053] Special state identification is used for specific hot standby scenarios, and is identified through a special state determination function, the expression of which is:

[0054]

[0055] in, For real-time voltage of equipment, For real-time current of the equipment, The no-load current threshold. , Taken from digital distribution network line model The rated parameters in;

[0056] Dynamic state update: The state is re-evaluated by recognizing changes in the operating mode. The expression is:

[0057]

[0058] in, This represents the change in operating mode. Threshold for determining changes in operating mode; when When necessary, the state fusion and discrimination process is automatically re-executed to update the device's comprehensive state database. The update cycle is synchronized with the power distribution automation system, with a default update cycle of 5 seconds per update.

[0059] As a preferred embodiment of the present invention, the specific steps of S2 are as follows:

[0060] S21. Core Instruction Requirements Analysis: An entity recognition model is constructed using a BERT-based bidirectional semantic parsing algorithm to accurately extract the core elements of the scheduling instructions, forming the instruction core vector:

[0061]

[0062] in, For the operational objective, For the scope of equipment involved, The BERT model uses a bidirectional Transformer structure to fuse contextual semantics, achieving an accuracy of at least 98% in recognizing scheduling terminology.

[0063] S22, Three-level hierarchical decomposition: Node-branch type topology generated based on S1 The node-branch relationships in the code are logically broken down into original instructions according to the hierarchical logic of region, bus, and equipment to obtain a set of basic operation instructions: ;

[0064] in, For circuit breaker opening and closing operations, For grounding wire disconnection / reconnection operations, each basic instruction contains only a single operation for a single device and clearly marks the device's unique ID;

[0065] S23. Disassembly rationality verification: Call the S1 device comprehensive status database. The real-time status of the device is verified using a status matching function to check the consistency between the command and the device status.

[0066]

[0067] in, For the target device involved in the instruction; exclude Conflicting instructions are used to form a stateless sequence of candidate instructions. , This represents the total number of candidate instructions.

[0068] As a preferred embodiment of the present invention, S3 specifically includes the following steps:

[0069] S31. Preliminary assessment of local area status: Delineate the directly associated area of ​​the equipment to be operated. The region is a set of devices with a topological distance ≤ 3 (including the device itself, directly connected devices, and adjacent devices). Abnormal instructions are filtered out using a local consistency check function. ;

[0070] in, This represents the total number of devices in a local area. The acceptable threshold; when If the system enters full network analysis, it will be returned to S2 for re-disassembly.

[0071] S32, In-depth analysis of the entire network topology:

[0072] Calculate the impact factor of the operation: Based on the power flow topology sensitivity algorithm, quantify the degree of impact of the operation on the overall network topology. ;

[0073] in, for The topological adjacency matrix at any given time. For nodes Importance This represents the total number of nodes in the entire network topology.

[0074] Two-level result fusion: Weighted integration of local verification and network-wide impact analysis results to form a comprehensive verification score: ;

[0075] in, The weights are for local analysis, with scores ranging from 0 to 1;

[0076] S33, Full-process verification:

[0077] Proposed ticket verification: Invokes the S1 intelligent anti-mistake reasoning algorithm based on the target operating mode. ,when , When the threshold for violation is reached, the instruction to be submitted for review is prohibited;

[0078] Verification and validation: A dual mechanism is employed, combining manual text review with Monte Carlo sampling simulation. Manual review verifies the standardization of instruction wording and the accuracy of equipment serial numbers; Monte Carlo simulation generates... A typical operating scenario, when ≥80% of the scenarios meet the requirements. and Approved in time The low-risk threshold;

[0079] Pre-execution verification: Update telemetry data With remote signaling data The error prevention reasoning and risk assessment algorithm will be re-executed, and if new risks are found, the algorithm will be returned for modification.

[0080] As a preferred embodiment of the present invention, S4 specifically includes the following steps:

[0081] S41. Real-time Risk Quantification: Calculates operational risk values ​​using real-time risk assessment algorithms.

[0082]

[0083] in: The total number of risk factors, including equipment failure risk, overload risk, personnel safety risk, and power supply reliability risk, with corresponding weights. The values ​​are 0.3, 0.25, 0.3, and 0.15 respectively, and ; For the first The state values ​​of risk factors are calculated from equipment operating parameters and topology status. For the first The dynamic risk coefficient of the risk factor is updated using the following formula: , As the benchmark coefficient, This is the sensitivity coefficient. for The absolute value of the deviation between the current state value and the reference value;

[0084] S42. Risk-based classification and control: Implement differentiated control measures based on risk values.

[0085]

[0086] Among them, operations involving malfunctions are prohibited from being executed, operations that have an impact require approval from the supervisor in charge, and operations that have no impact can be executed directly.

[0087] S43. Permission Compliance Verification: Verify operation eligibility through permission verification algorithms.

[0088]

[0089] in, For operator access levels, This represents the minimum privilege level required for the operation. For operation time, For the current duty period, To operate the equipment, For a set of authorized devices; only when Once the risk assessment is approved, the instruction enters the execution phase.

[0090] As a preferred embodiment of the present invention, S5 specifically includes the following steps:

[0091] S51, Command Issuance: Based on the connectivity of the S1 node branch topology, the command that has passed the S4 verification is issued to the distribution automation master station system in the topology association order of load side first, then power supply side, circuit breaker first, then disconnector.

[0092] S52. Operation completion verification:

[0093] Automatic verification: via state matching function The operation results are automatically verified using telemetry and teleindication data. The next step is prohibited at this time;

[0094] Manual verification: For disconnect switches, grounding switches, and other high-risk equipment without reliable remote signaling signals, operation can only continue after the status is manually confirmed in conjunction with video monitoring;

[0095] S53. Dynamic Updates and Secondary Error Prevention: After each operation is completed, the wiring change identification function is used. Trigger topology update, generate The updated real-time data from the entire network will be re-input into the S1 intelligent anti-misleading inference algorithm. Compared with the S4 risk assessment algorithm, when or When this occurs, subsequent commands are blocked and an audible and visual alarm is triggered.

[0096] As a preferred embodiment of the present invention, it further includes:

[0097] S61. Full-process data recording: Collect and store instruction dismantling results, verification records, risk quantification values, equipment status change data, operation anomaly information and handling results, and establish an optimized database;

[0098] S62, Core parameters are dynamically updated:

[0099] Adjusting the weight vector of the S1 intelligent error prevention inference algorithm This makes the error prevention judgment more closely aligned with actual operating scenarios; it also optimizes the equivalent impedance of the topology branches. Improve the accuracy of topology analysis; update the priority of the remote control operation safety constraint rule base. It adapts to changes in power distribution network structure and operational requirements; iteratively trains entity recognition models based on instruction samples in the optimized database to improve instruction parsing accuracy; and through continuous optimization, it forms a virtuous cycle of practice, optimization, and improvement, gradually enhancing the decomposition accuracy, verification reliability, and execution security of subsequent operations.

[0100] The beneficial technical effects of this invention are:

[0101] This invention achieves semantic-level parsing of comprehensive instructions through BERT entity recognition technology. It intelligently decomposes instructions into three levels: general instructions, sub-instructions, and operation items, achieving high accuracy and efficiency. It can generate standardized operation sequences without manual intervention, solving the problems of time-consuming and error-prone traditional manual decomposition. It is especially suitable for multi-device collaborative instruction processing in complex projects.

[0102] It adopts a dual constraint mechanism of local area topology analysis, network-wide topology verification, and full-process verification, and integrates multi-dimensional data such as device status, topology relationship, and anti-misoperation rules to effectively avoid the one-sidedness of local verification, improve the identification rate of illegal operations, and significantly reduce the risk of misoperation.

[0103] By dynamically quantifying and classifying risks and matching them with closed-loop access rights, time periods, and authorized devices, unauthorized operations are avoided while low-risk processes are simplified. Data accuracy is ensured by dynamic topology updates, and core parameters are iterated based on operational data to adapt to diverse scenarios. Coupled with a direct remote control mode with dual verification, the overall process time is significantly shortened, balancing the safety and efficiency of power distribution network scheduling. Attached Figure Description

[0104] Figure 1 This is a schematic diagram of the overall process of the present invention.

[0105] Figure 2 This is a schematic diagram of a node-branch type topology. Detailed Implementation

[0106] In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0107] Combination Figure 1-2 The present invention provides the following embodiments:

[0108] A method for intelligent decomposition and direct remote control of integrated instructions for power distribution network dispatching includes the following steps:

[0109] S1. Map the primary wiring structure of the physical distribution network lines to a digital distribution network line model using standardized object-oriented modeling tools; based on the digital model, call the topology analysis algorithm to generate a node-branch type topology structure containing topology nodes, topology branches, and topology adjacency matrices, which is automatically updated synchronously when the wiring changes; build a dual rule base including a remote control operation safety constraint rule base and a mis-reasoning rule base, and embed an intelligent mis-reasoning algorithm; deploy entity recognition technology, integrate the status variables of primary and secondary equipment and offline sign information, and identify the basic and special states of equipment through preset logic to form a dynamically updated comprehensive equipment status database;

[0110] S2. Analyze the core requirements of the scheduling integrated command through entity recognition technology. Based on the node branch topology, the basic operation command is disassembled into three levels: region, bus, and equipment. The equipment integrated status library is called to verify the matching between the disassembly command and the equipment status, eliminate conflicting commands, and form a candidate command sequence.

[0111] S3. Taking the candidate instruction sequence as input, first delineate the area directly associated with the device to be operated, and complete the initial judgment of the device status in the area based on the comprehensive device status database; then conduct in-depth analysis of the entire network topology through a digital model, and integrate the results of the two-level analysis; and then conduct full-process verification of drafting, review, and pre-execution.

[0112] S4. Input the topology analysis results of S3 and the equipment status data of S1 into the real-time risk assessment algorithm to quantify the risk of instructions and manage them in a hierarchical manner. With the risk classification results as a prerequisite, call the management rule error prevention logic of S1, and verify the operator's permission level, duty period and authorized equipment set through the permission verification algorithm. Only instructions with risk levels matching permissions are allowed to enter the execution stage.

[0113] S5. Issue instructions that have passed the S4 verification in the topology association order, and verify the completion of the operation through automatic telemetry and teleindication or video-assisted manual verification. After each operation, update the real-time data of the entire network and the node branch topology of S1 synchronously, and re-input the updated data into the intelligent anti-misoperation reasoning algorithm of S1 and the risk assessment algorithm of S4. If an anomaly is determined, block subsequent instructions and issue an alarm.

[0114] Furthermore, the specific steps for constructing the digital distribution network line model described in S1 are as follows:

[0115] Using a Public Information Modeling (CIM) standardized object-oriented modeling tool conforming to the IEC 61970 standard, physical distribution network primary equipment is transformed into structured digital instances through equipment attribute mapping functions. The expression of the mapping function is as follows: ;

[0116] in, For the first A complete set of attributes for each physical device, including device type, physical connection port location information, rated electrical parameters, and globally unique device number; The device is assigned a globally unique digital identifier, using a 16-bit string encoding rule. The first 8 bits are the administrative division code of the region to which the device belongs, and the last 8 bits are the device's sequential serial number. Equipment type is coded according to the following rules: 01=busbar, 02=circuit breaker, 03=disconnector, 04=load, 05=generator, 06=transformer, 07=grounding switch. The codes are unique and cannot be repeated. This is a set of three-dimensional coordinates of the electrical connection ports of the device, in the format of... The coordinate accuracy is 0.1m, which is used to accurately characterize the physical connection relationship between devices; This is a set of rated parameters for the equipment, including the rated voltage. Rated current Rated power ; The inherent impedance parameter of the equipment is calculated from the resistivity of the equipment material, its structural dimensions, and its temperature coefficient. The calculation formula is as follows: ,in The standard resistivity of the conductive material of the equipment, For the effective length of the conductive part, Cross-sectional area of ​​the conductive part For material temperature coefficient, This represents the difference between the actual ambient temperature and the standard temperature of 20℃.

[0117] Iterate through all primary devices in the distribution network, converting them one by one into digital device instances, and integrate them to form a digital distribution network line model: ;

[0118] in The total number of primary distribution network devices is represented by the digital distribution network line model, which is synchronized with the distribution automation system in real time with a data update cycle of no more than 1 second, thus realizing a one-to-one mapping between physical devices and digital models.

[0119] The CIM tool, compliant with IEC 61970, ensures compatibility between the digital model and industry-standard data, enabling cross-system data interoperability. Through equipment attribute mapping functions, core attributes of physical equipment, such as type, connection relationships, and rated parameters, are structurally transformed into digital instances, guaranteeing a one-to-one mapping between physical and digital data. Equipment coding rules ensure global uniqueness of equipment, avoiding identification conflicts. Inherent impedance calculation incorporates factors such as material, size, and temperature to accurately reflect the actual electrical characteristics of the equipment. The model is synchronized with the power distribution automation system in real time, ensuring that the digital model always closely matches the operating status of the physical equipment, providing accurate data support for subsequent topology analysis and command verification.

[0120] Furthermore, the specific steps for generating the node-branch type topology structure in the construction of the basic model and support system described in S1 are as follows:

[0121] Topology node extraction: Core topology nodes are filtered using a topology node extraction function. The function expression is:

[0122] ;

[0123] in, Refers to core equipment such as busbars, circuit breakers, disconnect switches, generators, and transformers. Refers to loads and grounding switch equipment; generates a set of topology nodes. , This represents the total number of topology nodes. It serves as a globally unique identifier for the node, and is associated with the device. There is an associated mapping relationship;

[0124] Topology branch extraction: Standardized branches are generated through topology branch extraction and isometry functions. The function expression is as follows:

[0125]

[0126] in, This serves as the unique identifier for the branch road. For branch equivalent impedance, single-port devices are calculated based on node-to-ground equivalent impedance. Two-port devices are calculated based on series or parallel impedance; series impedance is... , in parallel The three-port device is disassembled into three independent node-to-node branches and the equivalent impedance is calculated for each branch. Topology constraint clauses corresponding to branches, and remote operation safety constraint rule base. Match each rule in the table;

[0127] Construction of the topological adjacency matrix: The matrix expression is as follows:

[0128]

[0129] in, Represents a constrained, disconnected branch, with a value of To satisfy computational logic;

[0130] Dynamic topology update: Topology synchronization update is triggered by a wiring change identification function. The function expression is:

[0131]

[0132] in, The threshold for determining wiring changes is calculated based on the normalized change in equipment attributes; when When the time comes, the topology update process is automatically triggered, generating the updated topology structure. Ensure that the topology is consistent with the actual distribution network wiring.

[0133] The topology node extraction function focuses on key electrical nodes in the distribution network by filtering core equipment and devices with electrical connection breakpoints, reducing the impact of redundant nodes on analysis efficiency. Branch equivalent impedance employs differentiated calculation methods for different port devices to adapt to their electrical characteristics and ensure the accuracy of impedance parameters. The topology adjacency matrix uses specific impedance values ​​to represent allowed connected branches and infinity to represent prohibited connected branches, clearly depicting the electrical constraints between nodes. The wiring change identification function automatically triggers topology updates by comparing changes in device attributes with thresholds, ensuring that the topology structure is consistent with the actual distribution network wiring in real time and avoiding instruction processing based on distorted topology.

[0134] Furthermore, the specific steps for building the dual rule base in the construction of the basic model and supporting system are as follows:

[0135] Construction of a remote control operation safety constraint rule base: Hierarchical control of constraint rules is achieved through a rule priority quantification function, the function expression of which is:

[0136]

[0137] in:

[0138] For the weighting coefficients, satisfying The default value is the risk level weight. Historical trigger frequency weight Scope of influence weight It can be dynamically adjusted according to the needs of power distribution network operation and maintenance;

[0139] For the first Safety constraints The rules cover all scenarios of violations in distribution network operations, including prohibiting the closing of grounding switches while the circuit is energized and prohibiting the pulling of disconnect switches under load.

[0140] The value ranges from 0 to 1, representing the operational risk level corresponding to the rule. Specifically, the value is 1.0 for the highest risk level, 0.9 for the first level, 0.7 for the second level, and 0.5 for the third level.

[0141] The weight for the historical trigger frequency of the rule, with a value range of 0-1, is calculated as follows: Regular updates, such as monthly updates;

[0142] The weight of the impact range after the rule is triggered is 0-1, and is divided into levels according to the number of affected users: 1.0 for 100,000 users and above, 0.8 for 1.1 million users and 0.5 for less than 10,000 users.

[0143] Based on the above functions, constraint rules for all scenarios of distribution network operations are sorted out, forming a safety constraint rule library with priority identifiers. , To ensure the total number of security constraint rules, the rule base synchronously stores the topology node / branch association information corresponding to each rule;

[0144] Construction of the anti-misoperation reasoning rule base: It integrates five layers of anti-misoperation logic, including ontology state, relation state, device attributes, operation sequence, and safety distance, and embeds a fuzzy comprehensive evaluation-type intelligent anti-misoperation reasoning algorithm. The algorithm expression is as follows: ;

[0145] in: To prevent misjudgment of the result vector, For the operation belonging to the first The membership degree of the anti-misconduct category, with a value ranging from 0 to 1. To prevent accidental violations, the total number of categories includes violations related to the five prevention measures and high-risk misoperation violations, totaling 12 categories;

[0146] The weight vector of error-prevention factors is determined by the analytic hierarchy process (AHP) and satisfies the following conditions: The default value is the original state. Relationship status Device attributes Operation sequence Safe distance ;

[0147] It is a fuzzy relation matrix. For the first The error prevention factor for the first The membership degree of the anti-misconduct category, with a value range of 0-1, is jointly determined by historical fault data and industry expert experience, and is regularly verified and updated.

[0148] For fuzzy synthesis operators, a weighted average operator is used, and the specific calculation formula is as follows: This ensures a balanced integration of multiple error-prevention factors.

[0149] The priority quantification function of the remote control operation safety constraint rule base achieves differentiated control of rules through three weights: risk level, historical trigger frequency, and scope of impact, prioritizing the effectiveness of rules with high risk, high frequency of triggering, and wide-ranging impact. The error prevention reasoning rule base integrates five layers of logic, including ontology state and relation state, covering error prevention needs across all scenarios. In the fuzzy comprehensive evaluation algorithm, the weight vector highlights key error prevention factors, the fuzzy relation matrix quantifies the correlation strength between factors and violations, and the weighted average synthesis operator balances and integrates multi-dimensional information, avoiding the one-sidedness of single-factor judgment and improving the accuracy of error prevention judgment.

[0150] Furthermore, the construction of the comprehensive equipment status library in the basic model and support system construction described in S1 involves the following specific steps:

[0151] Multi-source state information fusion integrates state data from multiple heterogeneous devices through a state normalization function, the expression of which is: ;

[0152] in: To integrate the weighting coefficients, satisfy the following conditions: The default value is the weight of the primary device state quantity. Secondary equipment status quantity weights Offline signage information weight ;

[0153] For the device to be identified, and the digital distribution network line model The device instances in the database correspond one-to-one;

[0154] This is a primary equipment status variable, with a normalized value range of 0-1, defined as: closed state = 1, open state = 0, abnormal state = 0.5;

[0155] For secondary equipment status variables, the normalized value range is 0-1, defined as: protection device normal = 1, alarm status = 0.5, fault status = 0;

[0156] For offline signage information, the normalized value range is 0-1, defined as: Operation sign = 1, Maintenance sign = 0, Test sign = 0.3;

[0157] Device status determination includes basic status determination and special status determination. Basic status determination is implemented through a basic status mapping function, the expression of which is:

[0158]

[0159] Special state identification is used for specific hot standby scenarios, and is identified through a special state determination function, the expression of which is:

[0160]

[0161] in, For real-time voltage of equipment, For real-time current of the equipment, The no-load current threshold. , Taken from digital distribution network line model The rated parameters in;

[0162] Dynamic state update: The state is re-evaluated by recognizing changes in the operating mode. The expression is:

[0163]

[0164] in, This represents the change in operating mode. Threshold for determining changes in operating mode; when When necessary, the state fusion and discrimination process is automatically re-executed to update the device's comprehensive state database. The update cycle is synchronized with the power distribution automation system, with a default update cycle of 5 seconds per update.

[0165] Multi-source status information fusion employs weighted allocation, prioritizing primary equipment status quantities while also considering secondary equipment status and offline signage information to ensure comprehensive status identification. Status normalization maps different types of status quantities to the [0,1] interval, eliminating dimensional differences and facilitating unified judgment. Basic status is divided according to the normalized status value interval, conforming to the actual hierarchical level of distribution network equipment operation status. Special status is judged through voltage and current thresholds to accurately identify hot standby sub-scenarios. The operation mode change identification function automatically triggers status updates by comparing the status change quantity with the threshold, ensuring that the status database reflects the equipment operating conditions in real time.

[0166] Furthermore, the specific steps of S2 are as follows:

[0167] S21. Core Instruction Requirements Analysis: An entity recognition model is constructed using a BERT-based bidirectional semantic parsing algorithm to accurately extract the core elements of the scheduling instructions, forming the instruction core vector:

[0168]

[0169] in, For operational objectives, such as load transfer on the 10kV XX line, power outage for maintenance of the XX busbar, etc.; The scope of equipment involved includes the unique ID and type of the equipment; The operation time window is in the format YYYY-MM-DDHH:MM-HH:MM; the BERT model, through a bidirectional Transformer structure, fuses contextual semantics and achieves an accuracy of no less than 98% in recognizing scheduling terminology.

[0170] S22, Three-level hierarchical decomposition: Node-branch type topology generated based on S1 The node-branch relationships in the code are logically broken down into original instructions according to the hierarchical logic of region, bus, and equipment to obtain a set of basic operation instructions: ;

[0171] in, For circuit breaker opening and closing operations, For grounding wire disconnection / reconnection operations, each basic instruction contains only a single operation for a single device and clearly marks the device's unique ID;

[0172] S23. Disassembly rationality verification: Call the S1 device comprehensive status database. The real-time status of the device is verified using a status matching function to check the consistency between the command and the device status.

[0173]

[0174] in, For the target device involved in the instruction; exclude Conflicting instructions are used to form a stateless sequence of candidate instructions. , This represents the total number of candidate instructions.

[0175] The BERT-based entity recognition model resolves ambiguity in scheduling instructions by fusing contextual semantics through a bidirectional Transformer structure, enabling accurate extraction of core elements, including operation objectives, equipment scope, and time windows. It breaks down instructions into three levels: region, bus, and equipment, aligning with the distribution network topology hierarchy and transforming fuzzy comprehensive instructions into standardized basic operation instructions. A state matching function compares the decomposed instructions with the real-time status of the equipment's comprehensive status database, eliminating conflicting instructions and preventing erroneous operations caused by incompatibility between instructions and the current equipment status. This provides a conflict-free candidate instruction sequence for subsequent verification.

[0176] Furthermore, S3 specifically includes the following steps:

[0177] S31. Preliminary assessment of local area status: Delineate the directly associated area of ​​the equipment to be operated. The region is a set of devices with a topological distance ≤ 3, including the device itself, directly connected devices, and adjacent devices. Abnormal instructions are filtered out using a local consistency check function. ;

[0178] in, This represents the total number of devices in a local area. The acceptable threshold; when If the system enters full network analysis, it will be returned to S2 for re-disassembly.

[0179] S32, In-depth analysis of the entire network topology:

[0180] Calculate the impact factor of the operation: Based on the power flow topology sensitivity algorithm, quantify the degree of impact of the operation on the overall network topology. ;

[0181] in, for The topological adjacency matrix at any given time; For nodes The importance value ranges from 0 to 1, with hub nodes and heavily loaded nodes having higher values. This represents the total number of nodes in the entire network topology.

[0182] Two-level result fusion: Weighted integration of local verification and network-wide impact analysis results to form a comprehensive verification score: ;

[0183] in, The weights are for local analysis, with scores ranging from 0 to 1;

[0184] S33, Full-process verification:

[0185] Proposed ticket verification: Invokes the S1 intelligent anti-mistake reasoning algorithm based on the target operating mode. ,when , When the threshold for violation is reached, the instruction to be submitted for review is prohibited;

[0186] Verification and validation: A dual mechanism is employed, combining manual text review with Monte Carlo sampling simulation. Manual review verifies the standardization of instruction wording and the accuracy of equipment serial numbers; Monte Carlo simulation generates... A typical operating scenario, when ≥80% of the scenarios meet the requirements. and Approved in time The low-risk threshold;

[0187] Pre-execution verification: Update telemetry data With remote signaling data The error prevention reasoning and risk assessment algorithm will be re-executed, and if new risks are found, the algorithm will be returned for modification.

[0188] Local areas are defined as sets of devices with a topological distance ≤3, focusing on core devices related to operations. A local consistency check function is used to quickly filter out obviously abnormal commands, improving check efficiency. A comprehensive network topology analysis employs a power flow topology sensitivity algorithm to quantify the impact of operations on the entire network. Combined with node importance, it comprehensively assesses the global risk of operations. The fusion of two-level results prioritizes the reliability of local correlation checks while also considering the comprehensiveness of the overall network impact analysis. Full-process checks cover all stages of the operation. Proposed ticket checks prevent non-compliant commands from entering the review process. Review checks use Monte Carlo simulations to verify adaptability to multiple scenarios. Pre-execution checks are based on real-time data updates to ensure commands are adapted to the current power grid state.

[0189] Furthermore, S4 specifically includes the following steps:

[0190] S41. Real-time Risk Quantification: Calculates operational risk values ​​using real-time risk assessment algorithms.

[0191]

[0192] in: The total number of risk factors, including equipment failure risk, overload risk, personnel safety risk, and power supply reliability risk, with corresponding weights. The values ​​are 0.3, 0.25, 0.3, and 0.15 respectively, and ; For the first The state values ​​of risk factors are calculated from equipment operating parameters and topology status. For the first The dynamic risk coefficient of the risk factor is updated using the following formula: , This is a benchmark coefficient, set based on power safety regulations. This is the sensitivity coefficient. for The absolute value of the deviation between the current state value and the reference value;

[0193] S42. Risk-based classification and control: Implement differentiated control measures based on risk values.

[0194]

[0195] Among them, operations involving malfunctions are prohibited from being executed, operations that have an impact require approval from the supervisor in charge, and operations that have no impact can be executed directly.

[0196] S43. Permission Compliance Verification: Verify operation eligibility through permission verification algorithms.

[0197]

[0198] in, For operator access levels, This represents the minimum privilege level required for the operation. For operation time, For the current duty period, To operate the equipment, For a set of authorized devices; only when Once the risk assessment is approved, the instruction enters the execution phase.

[0199] The real-time risk assessment algorithm selects four core risk factors: equipment failure, overload, personnel safety, and power supply reliability. The weight allocation emphasizes the core position of personnel safety and equipment failure. The dynamic risk coefficient is dynamically adjusted through state deviation and sensitivity coefficient to adapt to the real-time operating status of the power grid. The risk classification threshold is set based on power safety regulations to achieve differentiated control of fault-related, impact-related, and non-impact-related operations. The authorization verification algorithm checks the operation qualification from three elements: authorization level, on-duty time period, and authorized equipment. It is linked with risk classification to ensure that high-risk operations can only be performed by authorized personnel, and low-risk operations simplify the process, balancing safety and efficiency.

[0200] Furthermore, S5 specifically includes the following steps:

[0201] S51, Command Issuance: Based on the connectivity of the S1 node branch topology, the command that has passed the S4 verification is issued to the distribution automation master station system in the topology association order of load side first, then power supply side, circuit breaker first, then disconnector.

[0202] S52. Operation completion verification:

[0203] Automatic verification: via state matching function The operation results are automatically verified using telemetry and teleindication data. The next step is prohibited at this time;

[0204] Manual verification: For disconnect switches, grounding switches, and other high-risk equipment without reliable remote signaling signals, operation can only continue after the status is manually confirmed in conjunction with video monitoring;

[0205] S53. Dynamic Updates and Secondary Error Prevention: After each operation is completed, the wiring change identification function is used. Trigger topology update, generate The updated real-time data from the entire network will be re-input into the S1 intelligent anti-misleading inference algorithm. Compared with the S4 risk assessment algorithm, when or When this occurs, subsequent commands are blocked and an audible and visual alarm is triggered.

[0206] The command issuance sequence is set according to "load side first, then power supply side; circuit breaker first, then disconnector," which complies with the distribution network operation safety regulations and avoids serious violations such as disconnecting disconnectors under load or closing grounding switches while energized. The operation completion is verified by a combination of automatic and manual verification. Automatic verification relies on telemetry and teleindication data to efficiently verify the status of routine equipment, while manual verification is used for high-risk equipment without reliable telemetry data. Video monitoring is used to supplement the verification to ensure that no part of the operation completion judgment is missed. After each operation, a topology update and secondary error prevention are triggered. The updated real-time data is used to re-execute error prevention reasoning and risk assessment to deal with potential risks caused by sudden changes in the power grid status during the operation. Abnormal commands are blocked and alarms are triggered to achieve dynamic safety control of the execution process.

[0207] Furthermore, it also includes:

[0208] S61. Full-process data recording: Collect and store instruction dismantling results, verification records, risk quantification values, equipment status change data, operation anomaly information and handling results, and establish an optimized database;

[0209] S62, Core parameters are dynamically updated:

[0210] Adjusting the weight vector of the S1 intelligent error prevention inference algorithm This makes the error prevention judgment more closely aligned with actual operating scenarios; it also optimizes the equivalent impedance of the topology branches. Improve the accuracy of topology analysis; update the priority of the remote control operation safety constraint rule base. It adapts to changes in power distribution network structure and operational requirements; iteratively trains entity recognition models based on instruction samples in the optimized database to improve instruction parsing accuracy; and through continuous optimization, it forms a virtuous cycle of practice, optimization, and improvement, gradually enhancing the decomposition accuracy, verification reliability, and execution security of subsequent operations.

[0211] The entire process of data recording and collection includes key data such as instruction disassembly, verification, risk, and equipment status, providing real-world sample support for model optimization. Through a closed-loop iteration of practical data, parameter updates, and performance improvement, the core parameters of the technical solution are continuously optimized, gradually improving disassembly accuracy, verification reliability, and execution security to meet the long-term operational needs of different power distribution network scenarios.

[0212] Taking a municipal distribution network dispatch center responsible for the operation and maintenance of the 10kV distribution network in the urban area as an example, due to the maintenance of the 10kV Chengdong Line, it is necessary to execute a comprehensive instruction to "transfer the load of transformer substations #1-#5 of the 10kV Chengdong Line to the 10kV Xincheng Line". The method of this invention is applied to complete the entire process operation, as follows:

[0213] Using a CIM modeling tool compliant with IEC 61970 standards, a digital distribution network model covering the Chengdong Line, Xincheng Line, and associated substations was constructed. This model automatically maps attribute information for 12 types of equipment, including circuit breakers and disconnect switches, generating a node-branch type topology with 32 topology nodes and 45 topology branches. A dual rule base was established, containing 89 safety constraint rules such as "prohibiting disconnect switches from being pulled under load," and embedding an intelligent anti-misoperation reasoning algorithm. Equipment remote signaling data, secondary protection status, and offline maintenance labeling information were integrated to form a dynamically updated comprehensive equipment status database, synchronized with distribution automation system data in real time.

[0214] BERT entity recognition technology was used to analyze the comprehensive instructions, extracting the operation target "load transfer", the equipment scope "Chengdong Line #1-#5 transformer areas + Xincheng Line associated switches", and the time window "09:00-10:00". These were then broken down into 6 basic operation instructions at three levels: "Area (Chengdong area) - Busbar (#2 busbar) - Equipment (circuit breaker / disconnector)". The equipment status database was called for verification, and it was found that the Chengdong Line #3 disconnector was in a "maintenance listed" state. Corresponding conflicting instructions were eliminated, resulting in a sequence of 5 candidate instructions.

[0215] A region with a topological distance of ≤3 between the devices to be operated was defined. After initial local state assessment, the states of 11 devices were found to be compatible and met the threshold requirements for full network analysis. The impact of the operation was quantified using a power flow topology sensitivity algorithm, and the results of local and full network verifications were combined, resulting in a comprehensive score of 0.85. The draft verification, review verification, and pre-execution verification were then completed sequentially.

[0216] The real-time risk assessment algorithm calculated a risk value of 0.62, classifying it as an impactful risk that requires approval from the supervisor in charge. The on-duty dispatcher's permission level was verified to be Level 3, matching the minimum permission required for the operation. Furthermore, the dispatcher was on duty during the specified time period, and the authorized equipment included the target line. Therefore, the instruction passed the screening.

[0217] Commands were issued sequentially from the load side to the power supply side. The completion of operations such as the opening of the Chengdong Line #1 circuit breaker and the closing of the Xincheng Line #2 circuit breaker was automatically verified through telemetry and teleindication. For the #4 disconnector switch, which lacked reliable telemetry, the status was manually confirmed in conjunction with video monitoring. After each operation, the topology was automatically updated, and anti-misoperation reasoning and risk assessment were re-executed. No abnormalities were found. The entire operation took 42 minutes, which is 58% shorter than the traditional manual mode.

[0218] Record the command decomposition results, verification records and equipment status change data, update the relationship status weight coefficients of the anti-misleading algorithm, optimize the calculation parameters of the equivalent impedance of the topology branch, and provide more accurate model support for subsequent similar load transfer commands.

[0219] No misoperation occurred in this application, and the power outage duration for users during the load transfer process was short, verifying the safety and efficiency of this invention in actual distribution network scheduling.

[0220] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for intelligent decomposition and direct remote control of integrated instructions for distribution network dispatching, characterized in that, Includes the following steps: S1. Map the primary wiring structure of the physical distribution network lines to a digital distribution network line model using standardized object-oriented modeling tools; based on the digital model, call the topology analysis algorithm to generate a node-branch type topology structure containing topology nodes, topology branches, and topology adjacency matrices, which is automatically updated synchronously when the wiring changes; build a dual rule base including a remote control operation safety constraint rule base and a mis-reasoning rule base, and embed an intelligent mis-reasoning algorithm; deploy entity recognition technology, integrate the status variables of primary and secondary equipment and offline sign information, and identify the basic and special states of equipment through preset logic to form a dynamically updated comprehensive equipment status database; S2. Analyze the core requirements of the scheduling integrated command through entity recognition technology. Based on the node branch topology, the basic operation command is disassembled into three levels: region, bus, and equipment. The equipment integrated status library is called to verify the matching between the disassembly command and the equipment status, eliminate conflicting commands, and form a candidate command sequence. S3. Taking the candidate instruction sequence as input, first delineate the area directly associated with the device to be operated, and complete the initial judgment of the device status in the area based on the comprehensive device status database; then conduct in-depth analysis of the entire network topology through a digital model, and integrate the results of the two-level analysis; and then conduct full-process verification of drafting, review, and pre-execution. S4. Input the topology analysis results of S3 and the equipment status data of S1 into the real-time risk assessment algorithm to quantify the risk of instructions and manage them in a hierarchical manner. With the risk classification results as a prerequisite, call the management rule error prevention logic of S1, and verify the operator's permission level, duty period and authorized equipment set through the permission verification algorithm. Only instructions with risk levels matching permissions are allowed to enter the execution stage. S5. Issue instructions that have passed the S4 verification in the topology association order, and verify the completion of the operation through automatic telemetry and teleindication or video-assisted manual verification. After each operation, update the real-time data of the entire network and the node branch topology of S1 synchronously, and re-input the updated data into the intelligent anti-misoperation reasoning algorithm of S1 and the risk assessment algorithm of S4. If an anomaly is determined, block subsequent instructions and issue an alarm.

2. The method for intelligent decomposition and direct remote control of integrated distribution network dispatching instructions according to claim 1, characterized in that, The specific steps for constructing the digital distribution network line model described in S1 are as follows: Using a Public Information Modeling (CIM) standardized object-oriented modeling tool conforming to the IEC 61970 standard, physical distribution network primary equipment is transformed into structured digital instances through equipment attribute mapping functions. The expression of the mapping function is as follows: ; in, For the first A complete set of attributes for each physical device, including device type, physical connection port location information, rated electrical parameters, and globally unique device number; The device is assigned a globally unique digital identifier, using a 16-bit string encoding rule. The first 8 bits are the administrative division code of the region to which the device belongs, and the last 8 bits are the device's sequential serial number. Equipment type is coded according to the following rules: 01=busbar, 02=circuit breaker, 03=disconnector, 04=load, 05=generator, 06=transformer, 07=grounding switch. The codes are unique and cannot be repeated. This is a set of three-dimensional coordinates of the electrical connection ports of the device, in the format of... The coordinate accuracy is 0.1m, which is used to accurately characterize the physical connection relationship between devices; This is a set of rated parameters for the equipment, including the rated voltage. Rated current Rated power ; The inherent impedance parameter of the equipment is calculated from the resistivity of the equipment material, its structural dimensions, and its temperature coefficient. The calculation formula is as follows: ,in The standard resistivity of the conductive material of the equipment, For the effective length of the conductive part, Cross-sectional area of ​​the conductive part For material temperature coefficient, This represents the difference between the actual ambient temperature and the standard temperature of 20℃. Iterate through all primary devices in the distribution network, converting them one by one into digital device instances, and integrate them to form a digital distribution network line model: ; in The total number of primary distribution network devices is represented by the digital distribution network line model, which is synchronized with the distribution automation system in real time with a data update cycle of no more than 1 second, thus realizing a one-to-one mapping between physical devices and digital models.

3. The method for intelligent decomposition and direct remote control of integrated distribution network dispatching commands according to claim 2, characterized in that, The specific steps for generating the node-branch type topology in the construction of the basic model and support system described in S1 are as follows: Topology node extraction: Core topology nodes are filtered using a topology node extraction function. The function expression is: ； in, Refers to core equipment such as busbars, circuit breakers, disconnect switches, generators, and transformers. Refers to loads and grounding switch equipment; generates a set of topology nodes. , This represents the total number of topology nodes. It serves as a globally unique identifier for the node, and is associated with the device. There is an associated mapping relationship; Topology branch extraction: Standardized branches are generated through topology branch extraction and isometry functions. The function expression is as follows: in, Used as the unique identifier for a branch road; For branch equivalent impedance, single-port devices are calculated based on node-to-ground equivalent impedance. Two-port devices are calculated based on series or parallel impedance; series impedance is... , in parallel The three-port device is disassembled into three independent node-to-node branches and the equivalent impedance is calculated for each branch. Topology constraint clauses corresponding to branches, and remote operation safety constraint rule base. Match each rule in the table; Construction of the topological adjacency matrix: The matrix expression is as follows: in, Represents a constrained, disconnected branch, with a value of To satisfy computational logic; Dynamic topology update: Topology synchronization update is triggered by a wiring change identification function. The function expression is: in, The threshold for determining wiring changes is calculated based on the normalized change in equipment attributes; when When the time comes, the topology update process is automatically triggered, generating the updated topology structure. Ensure that the topology is consistent with the actual distribution network wiring.

4. The method for intelligent decomposition and direct remote control of integrated distribution network dispatching commands according to claim 1, characterized in that, The specific steps for building the dual rule base in the construction of the basic model and supporting system are as follows: Construction of a remote control operation safety constraint rule base: Hierarchical control of constraint rules is achieved through a rule priority quantification function, the function expression of which is: in: For the weighting coefficients, satisfying The default value is the risk level weight. Historical trigger frequency weight Scope of influence weight It can be dynamically adjusted according to the needs of power distribution network operation and maintenance; For the first Safety constraints The rules cover all scenarios of violations in distribution network operations, including prohibiting the closing of grounding switches while the circuit is energized and prohibiting the pulling of disconnect switches under load. The value ranges from 0 to 1, representing the operational risk level corresponding to the rule. Specifically, the value is 1.0 for the highest risk level, 0.9 for the first level, 0.7 for the second level, and 0.5 for the third level. The weight for the historical trigger frequency of the rule, with a value range of 0-1, is calculated as follows: Regularly updated; The weight of the impact range after the rule is triggered is 0-1, and is divided into levels according to the number of affected users: 1.0 for 100,000 users and above, 0.8 for 1.1 million users and 0.5 for less than 10,000 users. Based on the above functions, constraint rules for all scenarios of distribution network operations are sorted out, forming a safety constraint rule library with priority identifiers. , To ensure the total number of security constraint rules, the rule base synchronously stores the topology node / branch association information corresponding to each rule; Construction of the anti-misoperation reasoning rule base: It integrates five layers of anti-misoperation logic, including ontology state, relation state, device attributes, operation sequence, and safety distance, and embeds a fuzzy comprehensive evaluation-type intelligent anti-misoperation reasoning algorithm. The algorithm expression is as follows: ; in: To prevent misjudgment of the result vector, For the operation belonging to the first The membership degree of the anti-misconduct category, with a value ranging from 0 to 1. To prevent accidental violations, the total number of violation categories; The weight vector of error-prevention factors is determined by the analytic hierarchy process (AHP) and satisfies the following conditions: The default value is the original state. Relationship status Device attributes Operation sequence Safe distance ; It is a fuzzy relation matrix. For the first The error prevention factor for the first The membership degree of the anti-misconduct category, with a value range of 0-1, is jointly determined by historical fault data and industry expert experience, and is regularly verified and updated. For fuzzy synthesis operators, a weighted average operator is used, and the specific calculation formula is as follows: This ensures a balanced integration of multiple error-prevention factors.

5. The intelligent decomposition and direct remote control method for integrated distribution network dispatching commands according to claim 4, characterized in that, The construction of the comprehensive equipment status library in the basic model and support system construction described in S1 involves the following specific steps: Multi-source state information fusion integrates state data from multiple heterogeneous devices through a state normalization function, the expression of which is: ; in: To integrate the weighting coefficients, satisfy the following conditions: The default value is the weight of the primary device state quantity. Secondary equipment status quantity weights Offline signage information weight ; For the device to be identified, and the digital distribution network line model The device instances in the database correspond one-to-one; This is a primary equipment status variable, with a normalized value range of 0-1, defined as: closed state = 1, open state = 0, abnormal state = 0.5; For secondary equipment status variables, the normalized value range is 0-1, defined as: protection device normal = 1, alarm status = 0.5, fault status = 0; For offline signage information, the normalized value range is 0-1, defined as: Operation sign = 1, Maintenance sign = 0, Test sign = 0.3; Device status determination includes basic status determination and special status determination. Basic status determination is implemented through a basic status mapping function, the expression of which is: Special state identification is used for specific hot standby scenarios, and is identified through a special state determination function, the expression of which is: in, For real-time voltage of equipment, For real-time current of the equipment, The no-load current threshold. , Taken from digital distribution network line model The rated parameters in; Dynamic state update: The state is re-evaluated by recognizing changes in the operating mode. The expression is: in, This represents the change in operating mode. Threshold for determining changes in operating mode; when When necessary, the state fusion and discrimination process is automatically re-executed to update the device's comprehensive state database. The update cycle is synchronized with the power distribution automation system, with a default update cycle of 5 seconds per update.

6. The method for intelligent decomposition and direct remote control of integrated distribution network dispatching instructions according to claim 1, characterized in that, The specific steps of S2 are as follows: S21. Core Instruction Requirements Analysis: An entity recognition model is constructed using a BERT-based bidirectional semantic parsing algorithm to accurately extract the core elements of the scheduling instructions, forming the instruction core vector: in, For the operational objective, For the scope of equipment involved, The BERT model uses a bidirectional Transformer structure to fuse contextual semantics, achieving an accuracy of at least 98% in recognizing scheduling terminology. S22, Three-level hierarchical decomposition: Node-branch type topology generated based on S1 The node-branch relationships in the code are logically broken down into original instructions according to the hierarchical logic of region, bus, and equipment to obtain a set of basic operation instructions: ; in, For circuit breaker opening and closing operations, For grounding wire disconnection / reconnection operations, each basic instruction contains only a single operation for a single device and clearly marks the device's unique ID; S23. Disassembly rationality verification: Call the S1 device comprehensive status database. The real-time status of the device is verified using a status matching function to check the consistency between the command and the device status. in, For the target device involved in the instruction; exclude Conflicting instructions are used to form a stateless sequence of candidate instructions. , This represents the total number of candidate instructions.

7. The method for intelligent decomposition and direct remote control of integrated distribution network dispatching commands according to claim 6, characterized in that, S3 specifically includes the following steps: S31. Preliminary assessment of local area status: Delineate the directly associated area of ​​the equipment to be operated. The region is a set of devices with a topological distance ≤ 3 (including the device itself, directly connected devices, and adjacent devices). Abnormal instructions are filtered out using a local consistency check function. ; in, This represents the total number of devices in a local area. The acceptable threshold; when If the system enters full network analysis, it will be returned to S2 for re-disassembly. S32, In-depth analysis of the entire network topology: Calculate the impact factor of the operation: Based on the power flow topology sensitivity algorithm, quantify the degree of impact of the operation on the overall network topology. ; in, for The topological adjacency matrix at any given time. For nodes Importance This represents the total number of nodes in the entire network topology. Two-level result fusion: Weighted integration of local verification and network-wide impact analysis results to form a comprehensive verification score: ; in, The weights are for local analysis, with scores ranging from 0 to 1; S33, Full-process verification: Proposed ticket verification: Invokes the S1 intelligent anti-mistake reasoning algorithm based on the target operating mode. ,when , When the threshold for violation is reached, the instruction to be submitted for review is prohibited; Verification and validation: A dual mechanism is employed, combining manual text review with Monte Carlo sampling simulation. Manual review verifies the standardization of instruction wording and the accuracy of equipment serial numbers; Monte Carlo simulation generates... A typical operating scenario, when ≥80% of the scenarios meet the requirements. and Approved in time The low-risk threshold; Pre-execution verification: Update telemetry data With remote signaling data The error prevention reasoning and risk assessment algorithm will be re-executed, and if new risks are found, the algorithm will be returned for modification.

8. The method for intelligent decomposition and direct remote control of integrated distribution network dispatching instructions according to claim 1, characterized in that, S4 specifically includes the following steps: S41. Real-time Risk Quantification: Calculates operational risk values ​​using real-time risk assessment algorithms. in: The total number of risk factors, including equipment failure risk, overload risk, personnel safety risk, and power supply reliability risk, with corresponding weights. The values ​​are 0.3, 0.25, 0.3, and 0.15 respectively, and ; For the first The state value (0~1) of risk factors is calculated from the equipment operating parameters and topology status; For the first The dynamic risk coefficient of the risk factor is updated using the following formula: , As the benchmark coefficient, This is the sensitivity coefficient. for The absolute value of the deviation between the current state value and the reference value; S42. Risk-based classification and control: Implement differentiated control measures based on risk values. Among them, operations involving malfunctions are prohibited from being executed, operations that have an impact require approval from the supervisor in charge, and operations that have no impact can be executed directly. S43. Permission Compliance Verification: Verify operation eligibility through permission verification algorithms. in, For operator access levels, This represents the minimum privilege level required for the operation. For operation time, For the current duty period, To operate the equipment, For a set of authorized devices; only when Once the risk assessment is approved, the instruction enters the execution phase.

9. The method for intelligent decomposition and direct remote control of integrated distribution network dispatching instructions according to claim 8, characterized in that, S5 specifically includes the following steps: S51, Command Issuance: Based on the connectivity of the S1 node branch topology, the command that has passed the S4 verification is issued to the distribution automation master station system in the topology association order of load side first, then power supply side, circuit breaker first, then disconnector. S52. Operation completion verification: Automatic verification: via state matching function The operation results are automatically verified using telemetry and teleindication data. The next step is prohibited at this time; Manual verification: For disconnect switches, grounding switches, and other high-risk equipment without reliable remote signaling signals, operation can only continue after the status is manually confirmed in conjunction with video monitoring; S53. Dynamic Updates and Secondary Error Prevention: After each operation is completed, the wiring change identification function is used. Trigger topology update, generate The updated real-time data from the entire network will be re-input into the S1 intelligent anti-misleading inference algorithm. Compared with the S4 risk assessment algorithm, when or When this occurs, subsequent commands are blocked and an audible and visual alarm is triggered.

10. The method for intelligent decomposition and direct remote control of integrated distribution network dispatching instructions according to claim 1, characterized in that, Also includes: S61. Full-process data recording: Collect and store instruction dismantling results, verification records, risk quantification values, equipment status change data, operation anomaly information and handling results, and establish an optimized database; S62, Core parameters are dynamically updated: Adjusting the weight vector of the S1 intelligent error prevention inference algorithm This makes the error prevention judgment more closely aligned with actual operating scenarios; it also optimizes the equivalent impedance of the topology branches. Improve the accuracy of topology analysis; update the priority of the remote control operation safety constraint rule base. It adapts to changes in power distribution network structure and operational requirements; iteratively trains entity recognition models based on instruction samples in the optimized database to improve instruction parsing accuracy; and through continuous optimization, it forms a virtuous cycle of practice, optimization, and improvement, gradually enhancing the decomposition accuracy, verification reliability, and execution security of subsequent operations.

Images