Load resource decision support method based on power grid topology and dynamic relationship database and related equipment

Abstract

The application belongs to the technical field of power system automation and load management, and discloses a load resource decision support method based on a power grid topology and a dynamic relationship database and related equipment; wherein the load resource decision support method based on the power grid topology and the dynamic relationship database comprises the following steps: inputting a regulation and control demand into a power grid topology snapshot model, outputting a most bottom layer feeder code set electrically connected with the regulation and control demand, inputting the most bottom layer feeder code set into a dynamic feeder-user relationship mapping database, outputting an original user list associated with the most bottom layer feeder code set, inputting the original user list into a rule engine and an intelligent optimization model, and outputting a structured load resource decision support list; the application can realize a second-level response, meet a minute-level timeliness requirement of a load regulation and control service, the frozen snapshot ensures that a matching benchmark under the same regulation and control target at any time is completely consistent and stable, and the application can also significantly improve the reliability and timeliness of feeder-user relationship data mapping.

Description

Technical Field

[0001] This invention relates to the field of power system automation and load management technology, specifically to a load resource decision support method and related equipment based on power grid topology and dynamic relational database. Background Technology

[0002] Currently, the construction of new power systems places higher demands on load management, urgently requiring rapid and accurate matching between grid-side control needs and user-side adjustable load resources. Load control operations involve scenarios such as demand response, orderly power consumption, and emergency control. Its core lies in accurately understanding the grid topology and feeder-user electrical connections to support the rapid development and decision-making of control schemes.

[0003] Existing technologies typically rely on real-time power grid topology calculations and feeder-user relationship data provided by the integrated operation and maintenance system for resource matching. However, this method has the following prominent problems: First, the power grid topology changes dynamically with the operating mode, and real-time topology calculations are too time-consuming (usually exceeding 2 hours), failing to meet the minute-level response time requirements. Furthermore, the calculation results are unstable due to topology fluctuations, leading to unreliable matching benchmarks. Second, feeder-user relationship data has blind spots and distortions, especially lacking effective coverage for 10kV and above dedicated line users. Moreover, data updates and maintenance are not closed-loop, and errors accumulate over time, easily resulting in missed or incorrect matching, seriously affecting the accuracy and fairness of control decisions.

[0004] Therefore, the existing technologies face problems such as excessively long real-time topology calculation time, unstable calculation results due to topology fluctuations, and blind spots and distortions in feeder-user relationship data, which urgently need to be addressed. Summary of the Invention

[0005] The purpose of this invention is to provide a load resource decision support method and related equipment based on power grid topology and dynamic relationship database to overcome the problems existing in the prior art. This invention can achieve second-level response and meet the minute-level timeliness requirements of load control business. At the same time, freezing snapshots ensures that the matching benchmark is completely consistent and stable at any time under the same control target. It can also significantly improve the reliability and timeliness of feeder-user relationship data mapping.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In a first aspect, the present invention provides a load resource decision support method based on power grid topology and dynamic relational database, comprising the following steps: The control requirements are obtained and input into the power grid topology snapshot model. The power grid topology snapshot model is based on the breadth-first traversal algorithm, which traverses the target equipment codes in the control requirements and outputs the set of the lowest-level feeder codes that are electrically connected to the control requirements. The lowest-level feeder code set is input into the dynamic feeder-user relationship mapping library. The dynamic feeder-user relationship mapping library is based on a hash index algorithm and performs matching queries using the lowest-level feeder code set as the key, outputting the original user list associated with the lowest-level feeder code set. The original user list is input into the rule engine and intelligent optimization model. The rule engine and intelligent optimization model filter, sort, and aggregate the original user list based on configurable filtering rules, multi-factor weighted scoring algorithms, and capacity constraint optimization algorithms, and output a structured load resource decision support list.

[0007] In some embodiments, obtaining the control requirements involves inputting the control requirements into a power grid topology snapshot model. This model, based on a breadth-first search algorithm, traverses the target device codes within the control requirements and outputs a set of the lowest-level feeder codes electrically connected to the control requirements. Specifically, this includes: The power grid topology snapshot model receives control requests and extracts the target equipment codes from the control requests. The target device code is converted into a standard code that matches the power grid one-map spatial database by using the built-in encoding mapping algorithm of the power grid topology snapshot model; Call the pre-frozen power grid topology snapshot, starting with the standard encoding, and perform a breadth-first traversal algorithm based on the hierarchical connection relationship fixed in the power grid topology snapshot. Search downstream along the electrical path layer by layer until the bottom feeder of all electrical connections is reached, and output the bottom feeder encoding set.

[0008] In some embodiments, the target equipment code includes the operation code of the target substation, and / or line, and / or transformer equipment; The power grid topology snapshot is a static electrical connection diagram generated during the previous day's peak load period; The hierarchical connection relationship fixed in the power grid topology snapshot is the station-line-transformer-feeder hierarchical connection relationship.

[0009] In some embodiments, the step of inputting the lowest-level feeder code set into a dynamic feeder-user relationship mapping library, wherein the dynamic feeder-user relationship mapping library is based on a hash index algorithm, performs matching queries using the lowest-level feeder code set as the key, and outputs the original user list associated with the lowest-level feeder code set, specifically includes: The dynamic feeder-user relationship mapping library receives the set of the lowest-level feeder codes and, based on the hash index algorithm, uses each lowest-level feeder code in the set of the lowest-level feeder codes as the query key to perform batch direct matching queries and obtain the matched users. Collect all matched users, remove duplicates, and output the original user list associated with the lowest-level feeder code set.

[0010] In some embodiments, after performing batch direct matching queries using each lowest-level feeder code as the query key to obtain the matched users, and before aggregating all the matched users, the method further includes: The number of users matched by each bottom-level feeder code is denoted as N. N ; Determine if NN is equal to 0. If it is, mark the current bottom-level feeder code as a zero-user feeder. If not, do not mark it. Determine if NN is greater than a preset threshold. If yes, mark the current bottom-level feeder code as an overloaded user feeder. If no, do not mark it. Zero-user feeders, excessive-user feeders, and matched users are marked as suspicious, and warning information is generated and pushed to the business personnel interface.

[0011] In some embodiments, the step of inputting the original user list into a rule engine and an intelligent optimization model, wherein the rule engine and intelligent optimization model, based on configurable filtering rules, multi-factor weighted scoring algorithms, and capacity constraint optimization algorithms, filter, sort, and aggregate the original user list, and output a structured load resource decision support list, specifically including: The rule engine and intelligent optimization model receive the original user list, configure multi-dimensional filtering conditions through a graphical interface, and after being compiled by the rule engine, perform second-level filtering on the original user list and output a set of candidate users. A multi-factor weighted scoring algorithm is used to calculate a comprehensive priority score for each candidate user in the candidate user set, generating a full-rank list. Based on the total control target, a greedy algorithm is used to select users sequentially from the full sorted list until the cumulative capacity meets the total control target, thus obtaining the optimal subset. The optimal subset is automatically grouped according to administrative region or power supply range. The adjustable capacity data contained in each group is counted. The user details, adjustable capacity data and total control indicators of each group are encapsulated into a structured load resource decision support list and then output.

[0012] In some embodiments, the multidimensional filtering criteria include resource type, reliability rating, and geographic region.

[0013] Secondly, the present invention provides a load resource decision support system based on power grid topology and dynamic relational database, comprising: The bottom-level feeder precise positioning module is used to obtain the control requirements and input the control requirements into the power grid topology snapshot model. The power grid topology snapshot model is based on the breadth-first traversal algorithm, which traverses the target equipment codes in the control requirements and outputs the set of bottom-level feeder codes that are electrically connected to the control requirements. The feeder-user matching module is used to input the lowest-level feeder code set into the dynamic feeder-user relationship mapping library. The dynamic feeder-user relationship mapping library is based on a hash index algorithm, uses the lowest-level feeder code set as the key to perform matching queries, and outputs the original user list associated with the lowest-level feeder code set. The rule-based intelligent filtering and decision output module is used to input the original user list into the rule engine and intelligent optimization model. The rule engine and intelligent optimization model filter, sort and aggregate the original user list based on configurable filtering rules, multi-factor weighted scoring algorithms and capacity constraint optimization algorithms, and output a structured load resource decision support list.

[0014] Thirdly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described load resource decision support method based on power grid topology and dynamic relational database.

[0015] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described load resource decision support method based on power grid topology and dynamic relational database.

[0016] The above technical solution has the following advantages or beneficial effects: Firstly, this invention provides a load resource decision support method based on power grid topology and a dynamic relational database. Compared with existing technologies, firstly, this invention uses a pre-frozen power grid topology snapshot instead of real-time dynamic topology as the calculation benchmark. Combined with a breadth-first traversal algorithm, it quickly maps control requirements to the lowest-level feeders, completely avoiding the performance bottleneck of real-time topology calculation taking more than 2 hours, achieving a second-level response, and meeting the minute-level timeliness requirements of load control services. At the same time, freezing the snapshot ensures that the matching benchmark is completely consistent and stable at any time under the same control objective, fundamentally solving the problem of unstable matching results caused by dynamic changes in power grid topology. Second, this invention constructs a dynamic feeder-user relationship mapping library independent of real-time topology. It achieves instant matching queries based on a hash index algorithm and incorporates an automatic identification and early warning mechanism for feeders with zero or excessive users, enabling timely detection and marking of suspicious relationships. Through a closed-loop feedback mechanism, omissions or mismatches discovered by business personnel during scheme preparation or execution can trigger reverse verification and alarms, driving continuous self-correction of the relationship library. This effectively overcomes the challenges of blind spots in operation and maintenance data (especially for 10kV and above dedicated line users) and error accumulation caused by non-closed-loop updates, significantly improving the reliability and timeliness of the mapped data. Third, this invention achieves second-level multi-dimensional automatic filtering through a configurable rule engine, replacing the inefficient traditional model relying on manual experience. It combines a multi-factor weighted scoring algorithm to comprehensively consider factors such as adjustable capacity, response reliability, and urgency of control to generate a full-scale ranked list. Then, a greedy algorithm is applied to quickly select the optimal user subset under capacity constraints. Finally, it automatically groups users by administrative region or power supply range and includes hierarchical statistical data, outputting a structured, ready-to-use load resource decision support list. This technology system effectively supports the rapid development of solutions for various business scenarios such as emergency control, orderly power consumption, and demand response, and significantly improves the automation level and decision-making efficiency of load management.

[0017] Secondly, this invention provides a load resource decision-making system based on power grid topology and a dynamic relational database. Through the collaborative work of three modules—precise positioning of underlying feeders, feeder-user matching, and rule-based intelligent screening—it achieves second-level response, highly reliable matching, and intelligent decision-making. By combining frozen topology snapshots with a dynamic relational database, it solves the industry problems of unstable matching benchmarks and distorted operation-distribution relationships, effectively supporting the rapid development of solutions for various load control scenarios and significantly improving the automation level and decision-making efficiency of load management operations.

[0018] Thirdly, the present invention provides a computer device that, through a processor executing a specific computer program, can efficiently implement the steps of the method of the present invention. When performing data processing tasks, the computer device can accurately perform numerical calculations and logical judgments, avoiding errors caused by human factors. At the same time, since the computer program has high stability and reliability, it can ensure the accuracy and consistency of the data processing results.

[0019] Fourthly, the present invention provides a computer-readable storage medium. By programming the steps of the method of the present invention into a computer program and storing it on the computer-readable storage medium, users can easily load these programs onto any compatible computer device and execute them without rewriting or converting the code, which greatly improves the convenience and flexibility of program execution. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the load resource decision support method based on power grid topology and dynamic relational database, as shown in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a computer device as shown in an embodiment of the present invention. Detailed Implementation

[0021] The present invention will be further described in detail below with reference to specific embodiments. These descriptions are for explanation purposes only and are not intended to limit the scope of the invention.

[0022] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] In the context of new power system construction, load management operations place extremely high demands on the rapid and accurate matching of grid-side control needs with user-side adjustable resources. Existing technologies typically rely on real-time grid topology calculations, statically or non-closed-loop maintained feeder-user relationship data, and human experience-driven resource selection models. When dealing with time-sensitive scenarios such as demand response, orderly power consumption, and emergency control, these technologies exhibit a series of prominent defects and shortcomings, primarily manifested in the following three aspects: First, dynamic changes in the power grid topology lead to unstable matching benchmarks and low computational efficiency. Existing methods, which locate the underlying feeders affected by control objectives, typically perform topology calculations based on real-time power grid models. However, the power grid topology dynamically adjusts with changes in operating conditions, and directly using real-time models as matching benchmarks carries the risk of inconsistent calculation results. More importantly, real-time topology calculations are too time-consuming; completing a full topology traversal in a large-scale power grid can take more than two hours, failing to meet the timeliness requirements for rapid load control scheme development (e.g., minute-level or ten-minute response), becoming a major bottleneck for operational response.

[0025] Secondly, the feeder-user relationship mapping suffers from data blind spots and timeliness distortion. Accurate matching hinges on mastering the precise electrical connection relationships between feeders and users. While existing technologies rely on data sources (such as the integrated operation and maintenance system) that provide the basic relationships, they have significant problems: First, data blind spots exist, especially for dedicated line users of 10kV and above, whose connection relationships with the power grid are often missing or outdated. Second, data update and maintenance are not closed-loop, lacking an effective feedback and correction mechanism, leading to the accumulation of errors in the relationship mapping database over time. When matching is performed based on flawed basic data, it is highly susceptible to missed matching (omitting users that should be regulated) or incorrect matching (including users that should not be regulated), severely impacting the accuracy and fairness of regulation decisions.

[0026] Finally, resource selection and decision-making output rely on manual processes, lacking intelligence and flexibility. After obtaining the initial user list, existing technologies typically require business personnel to manually select, sort, and group users based on experience to form a final executable solution. This process has three major drawbacks: First, it is inefficient and time-consuming, making it difficult to cope with sudden, large-scale control needs; second, it lacks unified and configurable standardized rules, and the selection logic may be inconsistent between different personnel or in different scenarios, resulting in poor comparability of decision results; third, it lacks sufficient intelligence, failing to comprehensively consider multiple dimensions such as user adjustable capacity, response reliability, geographical location, and reduction benefits for automatic optimization of sorting and grouping, making it difficult to quickly generate optimal resource combination solutions that meet specific capacity constraints and business objectives.

[0027] In summary, due to fundamental defects such as unstable benchmarks, inaccurate relationships, and unintelligent decision-making, existing technologies lead to difficulties in load control operations, including slow control response, high decision-making difficulty, and low efficiency in scheme preparation. There is an urgent need for a precise matching and decision support method that can ensure benchmark consistency, highly reliable relationships, and intelligent decision-making.

[0028] This invention provides a load resource decision support method based on power grid topology and a dynamic relational database. To systematically address the core pain points of existing technologies, such as unstable benchmarks, inaccurate relations, and unintelligent decision-making, this invention constructs a technical chain encompassing topology snapshot positioning, precise relational database matching, and intelligent rule engine filtering. (See also...) Figure 1 This includes the following steps: Step 1: Obtain the control requirements and input them into the power grid topology snapshot model. The power grid topology snapshot model is based on the breadth-first traversal algorithm, which traverses the target equipment codes in the control requirements and outputs the set of the lowest-level feeder codes that are electrically connected to the control requirements. Step 2: Input the lowest-level feeder code set into the dynamic feeder-user relationship mapping library. The dynamic feeder-user relationship mapping library is based on a hash index algorithm and performs matching queries with the lowest-level feeder code set as the key, outputting the original user list associated with the lowest-level feeder code set. Step 3: Input the original user list into the rule engine and intelligent optimization model. The rule engine and intelligent optimization model filter, sort and aggregate the original user list based on configurable filtering rules, multi-factor weighted scoring algorithms and capacity constraint optimization algorithms, and output a structured load resource decision support list.

[0029] This invention employs frozen topology snapshots to achieve second-level positioning, avoiding the bottleneck of real-time topology calculation; it constructs a dynamic feeder-user relationship database and a closed-loop feedback mechanism to solve the problems of blind spots and distortion in operation and maintenance data; and it combines a rule engine and multi-factor optimization algorithms to achieve intelligent filtering and grouping output. Overall, it improves the timeliness, accuracy, and automation level of load resource matching.

[0030] Example: This embodiment provides a load resource decision support method based on power grid topology and dynamic relational database, including the following steps: Step 1, Accurate positioning of the bottom feeders based on the frozen topology snapshot: Obtain the control requirements and input the control requirements into the power grid topology snapshot model. The power grid topology snapshot model is based on the breadth-first traversal algorithm to traverse the target equipment codes in the control requirements and output the set of bottom feeder codes that are electrically connected to the control requirements.

[0031] The core objective of this stage is to quickly and stably map regulation demands to the lowest-level feeders of the power grid, providing an accurate and efficient calculation benchmark for subsequent resource matching and avoiding the performance bottleneck of real-time topology calculation.

[0032] In some embodiments, step 1 specifically includes: Step 1.1: The power grid topology snapshot model receives the control requirements and extracts the target equipment codes from the control requirements.

[0033] In some embodiments, the target equipment code includes the operation code of the target substation, and / or line, and / or transformer equipment.

[0034] The system receives control requests from the dispatching side, which typically include the target substation, lines, transformers, and other equipment, along with their operating codes. Since equipment coding rules may differ between different systems, a coding mapping algorithm is first used to uniformly convert the equipment operating codes in the requests into a standard coding system that can be recognized and processed by the "Power Grid Map" spatial database. This step ensures the consistency of the data foundation for subsequent topology calculations.

[0035] Step 1.2: Using the encoding mapping algorithm built into the power grid topology snapshot model, the target device encoding is converted into a standard encoding that matches the power grid one-map spatial database.

[0036] Step 1.3: Call the pre-frozen power grid topology snapshot, starting from the standard code, and perform a breadth-first traversal algorithm based on the hierarchical connection relationship fixed in the power grid topology snapshot. Search downstream along the electrical path layer by layer until the bottom feeder of all electrical connections is reached, and output the bottom feeder code set.

[0037] In some embodiments, the power grid topology snapshot is a static electrical connection diagram generated during the peak load period of the previous day; the hierarchical connection relationship fixed in the power grid topology snapshot is a station-line-transformer-feeder hierarchical connection relationship.

[0038] This method abandons the calculation approach that relies on real-time dynamic topology, and instead uses a frozen grid topology snapshot generated during the previous day's peak load period as a unified, static calculation benchmark. This snapshot is a standardized snapshot of the grid's spatial connectivity at a specific moment, possessing stability. The algorithm of this invention starts with the standardized codes of the control target equipment, and based on the static electrical connections fixed in the snapshot, employs an automated breadth-first search (BFS) algorithm to search layer by layer down the path from station to line to transformer to feeder. This process quickly traverses and locates all the lowest-level feeders (including dedicated lines) electrically connected to the control target, and outputs the code set of these feeders. This scheme directly addresses the pain point of unstable benchmarks, transforming the original real-time topology calculation, which took more than 2 hours, into a second-level traversal based on static snapshots, meeting the requirements of minute-level business response.

[0039] In some embodiments, the training method for the power grid topology snapshot model includes the following steps: Load peak period data from multiple consecutive historical days are acquired from a single-map spatial database of the power grid. The static electrical connection relationships of each device are extracted daily to construct a time-series topology sample set. Consistency checks are performed on the time-series topology sample set, identifying and removing outliers caused by data anomalies or topology changes, resulting in a preprocessed time-series topology sample set. For each topology sample in the preprocessed time-series topology sample set, an adjacency matrix is ​​established using the target substation, line, and transformer equipment as nodes, according to the station-line-transformer-feeder hierarchy, serving as the topology structure feature. Using standard control requirements as test cases, a breadth-first traversal is performed on the adjacency matrix of each topology sample, calculating the traversal completeness and time consumption to obtain a comprehensive score. The topology sample with the highest comprehensive score is selected as the optimal topology snapshot. After freezing the optimal topology snapshot, an encoding mapping algorithm is established to map the target equipment codes of different systems to a standard encoding system, completing the training of the power grid topology snapshot model.

[0040] Step 2, High-reliability matching of feeder-user based on dynamic relational database: Input the lowest-level feeder code set into the dynamic feeder-user relational mapping database. The dynamic feeder-user relational mapping database is based on a hash index algorithm and performs matching queries with the lowest-level feeder code set as the key, outputting the original user list associated with the lowest-level feeder code set.

[0041] This phase aims to establish a high-precision, high-reliability mapping from the feeder list to specific adjustable user resources, resolving issues of missed or incorrect matching caused by inaccurate relationships.

[0042] In some embodiments, step 2 specifically includes: Step 2.1: The dynamic feeder-user relationship mapping library receives the set of lowest-level feeder codes. Based on the hash index algorithm, it performs batch direct matching queries using each lowest-level feeder code in the set as the query key to obtain the matched users.

[0043] Step 2.2: Obtain the number of users matched by each bottom-level feeder code, denoted as N. N ; Determine if NN is equal to 0. If yes, mark the current bottom-level feeder code as a zero-user feeder. If no, do not mark it. Determine if NN is greater than a preset threshold. If yes, mark the current bottom-level feeder code as an overloaded user feeder. If no, do not mark it. Mark the zero-user feeder, overloaded user feeder, and matched users as suspicious, generate warning information, and push it to the business personnel interface.

[0044] Step 2.3: Gather all matched users, remove duplicates, and output the original user list associated with the lowest-level feeder code set.

[0045] This invention constructs and maintains a dynamic relationship mapping library independent of real-time topology. The core data of this library comes from incremental updates synchronized periodically (e.g., weekly) by business systems such as operation and maintenance systems. The system automatically parses and refreshes the "feeder-user" relationship. A dedicated maintenance module is established to address blind spots and non-closed-loop update issues in 10kV and above leased line user data. The system can automatically monitor and prompt relationship changes, guiding business personnel to perform manual verification and confirmation. The key lies in the introduction of a closed-loop feedback mechanism: if business personnel discover missing or incorrect matching results during subsequent solution development or execution, they can submit feedback. This feedback will trigger reverse verification and data anomaly alarms in the system, thereby driving the relationship library to continuously self-correct and optimize, constantly improving data reliability and timeliness.

[0046] To efficiently process feeder lists, a fast matching algorithm based on a hash index built from feeder codes was designed. The algorithm directly queries the hash index of the relational database using the feeder code set output from the first stage, instantly returning the set of all associated users. Simultaneously, the algorithm incorporates data quality verification rules. For feeders with zero users (i.e., no users matched on a feeder) or excessive user feeders (an abnormally large number of users exceeding a reasonable range), the system automatically marks them as suspicious and issues a warning, prompting business personnel to pay close attention and intervene for verification, thereby enhancing the reliability of the final matching results.

[0047] In some embodiments, the training method for the dynamic feeder-user relationship mapping library includes the following steps: The system acquires incremental update data from the regularly synchronized operation and maintenance system, parses and refreshes the feeder-user relationship, and establishes a hash index using the feeder code as the key to form an initial mapping library. A dedicated maintenance module is established for 10kV and above dedicated line users to monitor changes in the power grid topology. When a missing or altered connection relationship is detected, a prompt message is generated to guide business personnel to manually verify and input the data, thus completing and correcting the initial mapping library. A closed-loop feedback mechanism is introduced to receive feedback information on missing or mismatched connections submitted by business personnel during scheme preparation or execution. Based on the feedback, a reverse verification process is triggered, alarms are issued for suspicious connection records, and data correction is driven. The correction results are updated to the initial mapping library, completing the iterative optimization of the mapping library and obtaining a dynamic feeder-user relationship mapping library.

[0048] Step 3, rule-based intelligent filtering and decision output for multiple scenarios: The original user list is input into the rule engine and intelligent optimization model. The rule engine and intelligent optimization model filter, sort and aggregate the original user list based on configurable filtering rules, multi-factor weighted scoring algorithms and capacity constraint optimization algorithms, and output a structured load resource decision support list.

[0049] This stage involves refining and intelligently processing the original user list obtained through matching to generate a structured decision support list that can be directly used for "solution development" business, thus solving the problem of "unintelligent decision-making".

[0050] In some embodiments, step 3 specifically includes: Step 3.1: The rule engine and intelligent optimization model receive the original user list, configure multi-dimensional filtering conditions through a graphical interface, and after compilation by the rule engine, perform second-level filtering on the original user list and output a set of candidate users.

[0051] Build a rule engine that supports flexible configuration. Business personnel can customize multi-dimensional and complex filtering conditions through a graphical interface or simple scripts, based on specific control scenarios (such as emergency control, orderly power consumption, and demand response). These conditions can cover multiple tag dimensions such as resource type, historical response reliability, current operating status, geographical region, and industry classification. After compiling the user-defined business rules, the rule engine quickly filters the massive user list in memory, achieving a return of a set of candidate users that meet specific business objectives within seconds, completely changing the inefficient traditional model of relying on manual experience to filter users one by one.

[0052] In some embodiments, multidimensional filtering criteria include resource type, reliability rating, and geographic region.

[0053] Step 3.2: Using a multi-factor weighted scoring algorithm, calculate the comprehensive priority score for each candidate user in the candidate user set and generate a full-rank list.

[0054] Step 3.3: Based on the total control index, apply a greedy algorithm to select users sequentially from the full sorted list until the cumulative capacity meets the total control index, thus obtaining the optimal subset.

[0055] The system performs intelligent sorting and aggregation optimization on the selected candidate user set. First, a multi-factor weighted scoring algorithm is used to comprehensively consider multiple influencing factors such as the user's adjustable capacity, the urgency of regulation, project affiliation, and the benefit of unit capacity reduction, calculating a comprehensive priority score for each user and generating a full ranking list. Then, business personnel can input specific total regulation targets (e.g., a 100MW load reduction). The system applies optimization algorithms (such as greedy algorithms) to quickly select the optimal subset of users that meet the core capacity constraints from the ranking list.

[0056] Step 3.4: Automatically group the optimal subset according to administrative region or power supply range, count the adjustable capacity data contained in each group, and encapsulate the user details, adjustable capacity data and total control indicators of each group into a structured load resource decision support list and output it.

[0057] The system outputs a structured resource decision list. This list supports automatic grouping of users by administrative region, power supply range, and other dimensions, and includes hierarchical statistical summary data of adjustable capacity, forming a ready-to-use modular decision support package that greatly improves the automation level and decision-making efficiency of load management scheme preparation.

[0058] In some embodiments, the training method for the rule engine and the intelligent optimization model includes the following steps: Collect user response data and business screening rules from historical load control scenarios as a training sample set, and extract resource type, reliability rating, and geographical region as label dimensions. Based on the training sample set, configure the set of screening conditions for the rule engine, and determine the priority and combination logic of each condition through expert experience or statistical analysis to complete the parameter training of the rule engine. Use the analytic hierarchy process or expert weighting method to train the weight parameters in the multi-factor weighted scoring model, so that the weight parameters fit the historical optimal decision results. Using the total control index as a constraint and the historical user ranking list as input, train the selection threshold and stopping condition of the greedy algorithm so that the optimal subset of the output satisfies the capacity constraint and maximizes the comprehensive score, thus completing the training of the rule engine and intelligent optimization model.

[0059] To make the technical solution of this invention clearer and more complete, the following detailed description, in conjunction with its technical route, provides a specific implementation method for a precise load resource matching and decision support method based on a frozen topology and dynamic relational database of a power grid. The specific implementation of this method mainly includes three stages, and its overall workflow follows the sequence of topology snapshot positioning, relational database matching, and intelligent filtering output.

[0060] In one embodiment of the present invention, a load resource decision support method based on power grid topology and dynamic relational database is provided, comprising the following steps: Step 1, Accurate positioning of the bottom feeders based on the frozen topology snapshot: Obtain the control requirements and input the control requirements into the power grid topology snapshot model. The power grid topology snapshot model is based on the breadth-first traversal algorithm to traverse the target equipment codes in the control requirements and output the set of bottom feeder codes that are electrically connected to the control requirements.

[0061] The core of this stage lies in using static, standardized grid topology snapshots to quickly and stably map control demands to the lowest-level feeder network, laying the foundation for subsequent matching.

[0062] Step 1.1, Control Demand Analysis and Coding Standardization: During implementation, the system receives control demand instructions from the dispatching side or other business systems. These instructions contain the target control equipment objects, such as specific substations, lines, or transformers. Since equipment coding rules may differ between different systems, this method first uses a built-in coding mapping algorithm to uniformly convert the equipment operation codes in the demand into a standard coding format completely consistent with the power grid spatial database, thus resolving the issue of data silos caused by inconsistent coding across multiple sources.

[0063] Step 1.2, Topology Snapshot Traversal and Feeder List Generation: The key to implementation lies in using a frozen static snapshot of the power grid topology as the calculation benchmark. This snapshot is preferably generated and fixed during the daily peak load period (such as the previous day's evening peak), representing a relatively stable and important operating state of the power grid. The algorithm starts with the standardized control target equipment code as the starting node, and automatically traverses the data based on the unchanging electrical connection relationships (station-line-transformer-feeder hierarchy and connection relationships) recorded in the snapshot using a breadth-first search (BFS) algorithm. The algorithm searches downstream layer by layer along the electrical connection path until it reaches all electrically connected bottom-level feeders (including dedicated lines), and outputs the standard code set of these feeders. This process avoids the bottleneck of excessively long processing time (usually exceeding 2 hours) caused by relying on real-time dynamic topology calculations, achieving a second-level response.

[0064] Step 2, High-reliability matching of feeder-user based on dynamic relational database: Input the lowest-level feeder code set into the dynamic feeder-user relational mapping database. The dynamic feeder-user relational mapping database is based on a hash index algorithm and performs matching queries with the lowest-level feeder code set as the key, outputting the original user list associated with the lowest-level feeder code set.

[0065] This phase aims to use an independent, high-quality relationship mapping library to accurately associate the feeder list obtained in the previous phase with specific user resources.

[0066] Step 2.1, Construction and Closed-Loop Maintenance of the Dynamic Relationship Database: During implementation, an independent feeder-user relationship mapping database needs to be built and maintained. Its core data source comes from incremental update data packets provided periodically (e.g., weekly) by the operation and maintenance system. The system automatically parses and refreshes the relationships within the database. For 10kV and above dedicated line users with insufficient operation and maintenance data coverage, a dedicated maintenance module is established. When the system detects changes in the power grid topology and the relationship database is not updated in a timely manner, it will automatically prompt business personnel to manually verify and enter the data. The innovative maintenance mechanism of this method lies in introducing closed-loop feedback: when business personnel find omissions or errors in the matching results during subsequent scheme preparation or execution, they can report this issue to the system. The system will trigger a reverse verification process based on the feedback, alerting and driving corrections for suspicious relationship records, thereby achieving continuous self-optimization and improved reliability of the relationship database.

[0067] Step 2.2, Fast Hash Matching and Suspicious Relationship Early Warning: To maximize speed during the matching process, a hash index-based query algorithm is designed. The system takes the feeder code set generated in the first stage as input, uses the feeder code as the hash key, and directly performs batch queries in the relational database, instantly returning the set of all associated users. Simultaneously, the algorithm embeds business verification rules to intelligently review the matching results. For example, it automatically identifies and marks empty feeders with zero associated users or excessively large feeders with an abnormally large number of associated users, issuing early warnings to business personnel as suspicious relationships, prompting them to pay close attention and verify, thereby further enhancing the reliability of the output results.

[0068] Step 3, rule-based intelligent filtering and decision output for multiple scenarios: The original user list is input into the rule engine and intelligent optimization model. The rule engine and intelligent optimization model filter, sort and aggregate the original user list based on configurable filtering rules, multi-factor weighted scoring algorithms and capacity constraint optimization algorithms, and output a structured load resource decision support list.

[0069] This stage involves in-depth processing of the matched raw user list to generate structured solutions that can be directly used for decision-making, based on the needs of different business scenarios.

[0070] Step 3.1, Configurable Rule Engine-Driven Automatic Filtering: The implementation provides a flexibly configurable rule engine. Business personnel can define multi-dimensional composite filtering conditions based on specific load control scenarios (such as power grid emergency control, orderly power consumption, demand response, etc.) through a graphical interface or by writing simple scripts. These conditions can cover resource type (such as industrial, commercial), historical response reliability rating, current equipment operating status, geographical region, voltage level, etc. After compiling the configured rules, the rule engine quickly filters the entire user set output from the second stage in memory, returning a set of candidate users that fully meet the current business objectives within seconds.

[0071] Step 3.2, Multi-factor Weighted Ranking and Optimized Grouping Output: The selected candidate user set is refined and organized. A multi-factor weighted scoring algorithm is used to comprehensively consider multiple factors such as each user's adjustable capacity, urgency of regulation, importance of project affiliation, and economic benefits of unit capacity reduction, to calculate a comprehensive priority score for each user and generate a full ranking list. Furthermore, if the regulation demand has a clear total capacity target (e.g., a 100MW load reduction), an optimization algorithm (e.g., a greedy algorithm) can be applied to quickly select the optimal subset of users that meet the capacity constraint from the ranking list. Finally, the system encapsulates the results into a structured decision package: the resource list can be automatically grouped by administrative region, power supply branch, etc., and includes hierarchical statistical data on adjustable capacity within each group, forming a ready-to-use, modularly combinable decision support document that greatly improves the efficiency of scheme preparation.

[0072] This invention provides a load resource decision support method based on power grid topology and dynamic relational database. Compared with the core pain points of existing technologies, such as unstable benchmarks, inaccurate relations, and unintelligent decision-making, the advantages of this invention are specifically reflected in the following aspects: First, this invention solves the problem of dynamic fluctuations in the matching benchmark, achieving precise positioning with millisecond-level consistency. Existing technologies rely on real-time topology calculations, which take more than two hours, and the results fluctuate frequently with changes in grid operation, leading to unstable matching benchmarks. This invention creatively uses a frozen grid topology snapshot generated during the previous day's peak load period as a unified, static calculation benchmark. Through an automated breadth-first traversal algorithm, control requirements (target stations, lines, and transformers) are quickly and directly mapped to the lowest-level feeders. This method completely avoids the performance bottleneck of real-time topology calculations, achieving a leap from "hourly" to "second-level" accuracy. It also ensures that at any given time, for the same control objective, the grid structure benchmark upon which the matching calculation relies is completely consistent and stable, fundamentally solving the problem of inconsistent matching results caused by real-time changes in grid topology, and providing a solid and reliable foundation for subsequent precise matching.

[0073] Secondly, this invention overcomes the industry challenge of inaccurate feeder-user relationship data by constructing a self-evolving and highly reliable feeder-user mapping system. Traditional methods rely on integrated feeder-user data, which suffers from untimely updates, data blind spots (such as dedicated line users), and a lack of error correction mechanisms, resulting in numerous omissions or mismatches in the feeder-user relationship mapping. This invention establishes an independently maintained and dynamically updated feeder-user relationship mapping library and introduces a closed-loop feedback and automatic verification mechanism to achieve continuous optimization and reliable matching of relationships. Specifically, it sets up a dedicated maintenance module for 10kV and above dedicated line user data, proactively identifies and alerts to data blind spots, guides manual verification, and compensates for the inherent defects of existing data. The key innovation lies in the closed-loop feedback mechanism: feedback from business personnel on actual matching results (such as discovering omissions or mismatches) can trigger the system's reverse verification and alarms, driving the relationship library to continuously self-correct and improve data quality. The algorithm uses a fast matching algorithm based on hash indexes to complete the mapping instantly, and has built-in rules to automatically warn of suspicious relationships such as zero-user feeders or excessive-user feeders, further enhancing the reliability of the output and the trust of business personnel.

[0074] Finally, this invention overturns the inefficient traditional manual screening model, providing structured and customizable intelligent decision support. Traditional load resource screening relies heavily on human experience, resulting in low efficiency, inconsistent rules, and difficulty in meeting the rapid decision-making needs under multiple scenarios and complex constraints. This invention, by constructing a configurable rule engine and a multi-factor intelligent optimization model, achieves automated and intelligent generation from the original user list to a structured decision solution. Business personnel can flexibly define multi-dimensional composite screening rules (such as resource type, response reliability, geographical region, etc.) through a graphical interface based on specific scenarios such as emergency control and orderly power consumption. The rule engine compiles and executes in seconds, quickly filtering out a set of candidate users that meet business objectives. Furthermore, a multi-factor weighted scoring algorithm is adopted to automatically calculate the comprehensive priority based on multiple factors such as user adjustable capacity, urgency, and reduction benefits, generating a full-rank list. Combined with the total control constraint, optimization algorithms (such as greedy algorithms) are applied to quickly select the optimal subset of users that meet the core capacity requirements, and automatic grouping by administrative region, power supply range, etc. is supported, outputting a modular decision support package with hierarchical statistical data. This technology system effectively supports the rapid development of solutions for various business scenarios such as demand response, orderly power consumption, and emergency control, and significantly improves the automation level and decision-making efficiency of load management.

[0075] In one embodiment of the present invention, a load resource decision support system based on power grid topology and dynamic relational database is provided, comprising: The bottom-level feeder precise positioning module is used to obtain the control requirements and input the control requirements into the power grid topology snapshot model. The power grid topology snapshot model is based on the breadth-first traversal algorithm, which traverses the target equipment codes in the control requirements and outputs the set of bottom-level feeder codes that are electrically connected to the control requirements. The feeder-user matching module is used to input the lowest-level feeder code set into the dynamic feeder-user relationship mapping library. The dynamic feeder-user relationship mapping library is based on a hash index algorithm, uses the lowest-level feeder code set as the key to perform matching queries, and outputs the original user list associated with the lowest-level feeder code set. The rule-based intelligent filtering and decision output module is used to input the original user list into the rule engine and intelligent optimization model. The rule engine and intelligent optimization model filter, sort and aggregate the original user list based on configurable filtering rules, multi-factor weighted scoring algorithms and capacity constraint optimization algorithms, and output a structured load resource decision support list.

[0076] See Figure 2 This embodiment provides a computer device, which includes a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor is the computing and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the computer storage medium to realize the corresponding method flow or corresponding function. The processor described in this embodiment can be used to execute related operations of a load resource decision support method based on power grid topology and dynamic relationship database.

[0077] This embodiment provides a computer-readable storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space, which stores the terminal's operating system; and, in this storage space, it also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM (Random Access Memory) or non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the load resource decision support method based on power grid topology and dynamic relationship database in this embodiment.

[0078] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0079] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0080] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0081] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A load resource decision support method based on power grid topology and dynamic relationship database, characterized in that, Includes the following steps: The control requirements are obtained and input into the power grid topology snapshot model. The power grid topology snapshot model is based on the breadth-first traversal algorithm, which traverses the target equipment codes in the control requirements and outputs the set of the lowest-level feeder codes that are electrically connected to the control requirements. The lowest-level feeder code set is input into the dynamic feeder-user relationship mapping library. The dynamic feeder-user relationship mapping library is based on a hash index algorithm and performs matching queries using the lowest-level feeder code set as the key, outputting the original user list associated with the lowest-level feeder code set. The original user list is input into the rule engine and intelligent optimization model. The rule engine and intelligent optimization model filter, sort, and aggregate the original user list based on configurable filtering rules, multi-factor weighted scoring algorithms, and capacity constraint optimization algorithms, and output a structured load resource decision support list.

2. The grid topology and dynamic relationship library based load resource decision support method according to claim 1, characterized in that, The process of obtaining control requirements involves inputting these requirements into a power grid topology snapshot model. This model, based on a breadth-first search algorithm, traverses the target equipment codes within the control requirements and outputs a set of the lowest-level feeder codes electrically connected to the control requirements. Specifically, this includes: The power grid topology snapshot model receives control requests and extracts the target equipment codes from the control requests. The target device code is converted into a standard code that matches the power grid one-map spatial database by using the built-in encoding mapping algorithm of the power grid topology snapshot model; Call the pre-frozen power grid topology snapshot, starting with the standard encoding, and perform a breadth-first traversal algorithm based on the hierarchical connection relationship fixed in the power grid topology snapshot. Search downstream along the electrical path layer by layer until the bottom feeder of all electrical connections is reached, and output the bottom feeder encoding set.

3. The grid topology and dynamic relationship library based load resource decision support method of claim 2, wherein, The target equipment code includes the operation code of the target substation, and / or line, and / or transformer equipment; The power grid topology snapshot is a static electrical connection diagram generated during the previous day's peak load period; The hierarchical connection relationship fixed in the power grid topology snapshot is the station-line-transformer-feeder hierarchical connection relationship.

4. The grid topology and dynamic relationship library based load resource decision support method of claim 1, wherein, The step of inputting the lowest-level feeder code set into the dynamic feeder-user relationship mapping library, which is based on a hash index algorithm and performs matching queries using the lowest-level feeder code set as the key, outputs the original user list associated with the lowest-level feeder code set, specifically including: The dynamic feeder-user relationship mapping library receives the set of the lowest-level feeder codes and, based on the hash index algorithm, uses each lowest-level feeder code in the set of the lowest-level feeder codes as the query key to perform batch direct matching queries and obtain the matched users. Collect all matched users, remove duplicates, and output the original user list associated with the lowest-level feeder code set.

5. The grid topology and dynamic relationship library based load resource decision support method of claim 4, wherein, After performing batch direct matching queries using the code of each lowest-level feeder as the query key to obtain the matched users, and before aggregating all the matched users, the process further includes: Obtain the number of users matched to each bottommost layer feeder code, denoted as N N ; Determine if NN is equal to 0. If it is, mark the current bottom-level feeder code as a zero-user feeder. If not, do not mark it. Determine if NN is greater than a preset threshold. If yes, mark the current bottom-level feeder code as an overloaded user feeder. If no, do not mark it. Zero-user feeders, excessive-user feeders, and matched users are marked as suspicious, and warning information is generated and pushed to the business personnel interface.

6. The grid topology and dynamic relationship library based load resource decision support method of claim 1, wherein, The original user list is input into the rule engine and intelligent optimization model. Based on configurable filtering rules, multi-factor weighted scoring algorithms, and capacity constraint optimization algorithms, the rule engine and intelligent optimization model filter, sort, and aggregate the original user list, outputting a structured load resource decision support list, specifically including: The rule engine and intelligent optimization model receive the original user list, configure multi-dimensional filtering conditions through a graphical interface, and after being compiled by the rule engine, perform second-level filtering on the original user list and output a set of candidate users. A multi-factor weighted scoring algorithm is used to calculate a comprehensive priority score for each candidate user in the candidate user set, generating a full-rank list. Based on the total control target, a greedy algorithm is used to select users sequentially from the full sorted list until the cumulative capacity meets the total control target, thus obtaining the optimal subset. The optimal subset is automatically grouped according to administrative region or power supply range. The adjustable capacity data contained in each group is counted. The user details, adjustable capacity data and total control indicators of each group are encapsulated into a structured load resource decision support list and then output.

7. The grid topology and dynamic relationship library based load resource decision support method according to claim 6, characterized in that, The multidimensional filtering criteria include resource type, reliability rating, and geographical region.

8. A load resource decision support system based on power grid topology and dynamic relational database, characterized in that, include: The bottom-level feeder precise positioning module is used to obtain the control requirements and input the control requirements into the power grid topology snapshot model. The power grid topology snapshot model is based on the breadth-first traversal algorithm, which traverses the target equipment codes in the control requirements and outputs the set of bottom-level feeder codes that are electrically connected to the control requirements. The feeder-user matching module is used to input the lowest-level feeder code set into the dynamic feeder-user relationship mapping library. The dynamic feeder-user relationship mapping library is based on a hash index algorithm, uses the lowest-level feeder code set as the key to perform matching queries, and outputs the original user list associated with the lowest-level feeder code set. The rule-based intelligent filtering and decision output module is used to input the original user list into the rule engine and intelligent optimization model. The rule engine and intelligent optimization model filter, sort and aggregate the original user list based on configurable filtering rules, multi-factor weighted scoring algorithms and capacity constraint optimization algorithms, and output a structured load resource decision support list.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the load resource decision support method based on power grid topology and dynamic relational database as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the load resource decision support method based on power grid topology and dynamic relational database as described in any one of claims 1-7.

Images