Power distribution system load transfer strategy evaluation method oriented to expert preference alignment

Abstract

The invention provides an expert preference alignment-oriented power distribution system load transfer strategy evaluation method, device and equipment and a medium, and relates to the technical field of power distribution system faults, an evaluation model training set is constructed, an evaluation standard candidate set is constructed through a preset large language model and experts, and the power distribution system load transfer strategy is evaluated based on the evaluation standard candidate set. Obtaining an evaluation model assertion set through a large language model, and obtaining an expert assertion set formed after an expert performs artificial binarization evaluation on the evaluation model training set; based on the consistency measurement of the expert assertion set and the evaluation model assertion set, screening out an optimal evaluation standard from the evaluation standard candidate set; and evaluating the load transfer strategy to be evaluated through the optimal evaluation standard. According to the method, the problems of subjectivity and low efficiency of a traditional evaluation method are solved, automation and intellectualization of the evaluation process are achieved, the accuracy, objectivity and reliability of strategy evaluation are improved, and powerful support is provided for safe and stable operation of a power distribution system.

Description

Technical Field

[0001] The present invention relates to the technical field of power distribution system failure, and in particular to a method, device, equipment and medium for evaluating a load transfer strategy of a power distribution system for expert preference alignment. Background Art

[0002] With the growth of economy and the improvement of quality of life, the stable operation of power system is particularly critical. Research on distribution network fault recovery and power supply reliability improvement can not only help to quickly restore power supply and reduce power outage losses, but also prevent faults by optimizing maintenance plans. However, the current methods are too dependent on the experience of operators and lack economic and reliability guarantees. The combination of power grid technology and large language model technology has become the main trend of future power grid development. However, how to effectively and accurately evaluate the output quality of these models has become a new problem. The core of evaluating large language model technology lies in the subjectivity of standard setting, the complexity of evaluation and the interpretability of results. We hope to build an efficient and reliable evaluation system by combining a hybrid active method of automation and manual verification, while paying attention to and effectively managing the "standard drift" phenomenon in the evaluation process to ensure the long-term adaptability and accuracy of the evaluation tool. Therefore, it is urgent to design an evaluation tool to help experts generate and evaluate the output of large language model technology.

[0003] The application of machine learning algorithms in the evaluation of load transfer strategies in distribution systems is mainly based on data-driven methods, using historical data to predict and evaluate the effects of different load transfer strategies. The basic idea is: first, collect relevant data from the distribution system, including load data, equipment status, and meteorological conditions; process and transform the collected data to extract representative features. Features may include historical load curves, peak power consumption times, seasonal characteristics, etc., in order to better capture the impact of load transfer strategies. However, traditional machine learning algorithms have difficulty using dispatching regulations in the power sector and are prone to falling into local optimality. Summary of the invention

[0004] The purpose of the present invention is to provide a method, device, equipment and medium for evaluating the load transfer strategy of a distribution system for expert preference alignment, which is used to solve the problems of difficulty in quantifying expert experience, low evaluation efficiency and strong subjectivity in the load transfer strategy evaluation process of the prior art, realize automated and intelligent load transfer strategy evaluation, and can accurately and efficiently evaluate the load transfer strategy and achieve a high degree of consistency with expert preferences.

[0005] In order to achieve the above object, the present invention provides a method for evaluating a load transfer strategy of a distribution system based on expert preference alignment, comprising: The load transfer strategy to be evaluated interacts with the simulation environment to generate a fault recovery result report, and together with the corresponding load transfer strategy, constructs an evaluation model training set; The semantic understanding of the regulatory text is performed through a preset large language model to generate multiple textual evaluation criteria, which together with the textual evaluation criteria provided by experts form a candidate set of evaluation criteria; Based on the candidate set of evaluation criteria, the evaluation model training set is binarized and evaluated through a large language model to obtain the evaluation model assertion set; Obtaining an expert assertion set formed by the experts performing manual binary evaluation on the evaluation model training set; Based on the consistency measurement between the expert assertion set and the evaluation model assertion set, the optimal evaluation criterion is selected from the evaluation criterion candidate set; The load transfer strategy to be evaluated is evaluated using the optimal evaluation criteria.

[0006] According to a distribution system load transfer strategy evaluation method for expert preference alignment provided by the present invention, the load transfer strategy to be evaluated interacts with the simulation environment, generates a fault recovery result report, and constructs an evaluation model training set together with the corresponding load transfer strategy, including: Construct fault scenarios in the simulation environment, load and execute the load transfer strategy to be evaluated; obtain the operating data, fault recovery process and performance evaluation data of the distribution system to form a fault recovery result report; and construct the evaluation model training set together with the fault recovery result report and the corresponding load transfer strategy.

[0007] According to a distribution system load transfer strategy evaluation method for expert preference alignment provided by the present invention, the control procedure text includes the operation procedures, fault handling procedures, safety specifications, and technical standards of the distribution system.

[0008] According to a distribution system load transfer strategy evaluation method for expert preference alignment provided by the present invention, based on the candidate set of evaluation criteria, the evaluation model training set is binarized and evaluated by a large language model to obtain an evaluation model assertion set, including: For each evaluation criterion in the evaluation criterion candidate set, each load transfer strategy and the corresponding fault recovery result report in the evaluation model training set are input into the large language model; Use the large language model to determine whether the load transfer strategy meets the evaluation criteria, and output a binary result of compliance or non-compliance; The binarization results of all load transfer strategies under all evaluation criteria are summarized to form an evaluation model assertion set.

[0009] According to a distribution system load transfer strategy evaluation method for expert preference alignment provided by the present invention, based on the consistency measurement of the expert assertion set and the evaluation model assertion set, the optimal evaluation criterion is screened out from the evaluation criterion candidate set, including: Statistical methods are used to calculate the coverage and false rejection rate of the evaluation model assertion set and the expert assertion set; The consistency index between the assertion set of the evaluation model and the assertion set of the expert is obtained based on the coverage rate and the false rejection rate; The candidate sets of evaluation criteria are sorted according to the consistency index, and the evaluation criterion with the highest consistency index is selected as the optimal evaluation criterion.

[0010] According to a distribution system load transfer strategy evaluation method for expert preference alignment provided by the present invention, the expression of the consistency index is:

[0011]

[0012] In the formula, Represents a set of assertions; for F The coverage rate, for F The false rejection rate, for F Consistency indicators; is the indicator function; if the load transfer strategy , the expert believes that it does not meet the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it does not meet the evaluation criteria provided by the experts, then =1; if the load transfer strategy , the expert believes that it meets the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it meets the evaluation criteria provided by the experts, then =1.

[0013] In a second aspect, the present invention provides a distribution system load transfer strategy evaluation device for expert preference alignment, comprising: A simulation interaction unit is used to interact the load transfer strategy to be evaluated with the simulation environment, generate a fault recovery result report, and jointly construct an evaluation model training set with the corresponding load transfer strategy; An evaluation criteria generation unit, which is used to perform semantic understanding of the regulatory procedure text through a preset large language model, generate multiple evaluation criteria in text form, and form an evaluation criteria candidate set together with the evaluation criteria in text form provided by experts; A first acquisition unit is used to perform binary evaluation on the evaluation model training set through a large language model based on the evaluation standard candidate set to obtain an evaluation model assertion set; The second acquisition unit is used to acquire an expert assertion set formed after the expert performs manual binary evaluation on the evaluation model training set; A screening unit, used for screening out the best evaluation criteria from the evaluation criteria candidate set based on the consistency measurement between the expert assertion set and the evaluation model assertion set; The evaluation unit is used to evaluate the load transfer strategy to be evaluated through the optimal evaluation standard.

[0014] According to a distribution system load transfer strategy evaluation device for expert preference alignment provided by the present invention, the simulation interaction unit is specifically used for: Construct fault scenarios in the simulation environment, load and execute the load transfer strategy to be evaluated; obtain the operating data, fault recovery process and performance evaluation data of the distribution system to form a fault recovery result report; and construct the evaluation model training set together with the fault recovery result report and the corresponding load transfer strategy.

[0015] According to a distribution system load transfer strategy evaluation device for expert preference alignment provided by the present invention, the control procedure text includes the operation procedure, fault handling process, safety specifications and technical standards of the distribution system.

[0016] According to a distribution system load transfer strategy evaluation device for expert preference alignment provided by the present invention, the first acquisition unit is specifically used for: For each evaluation criterion in the evaluation criterion candidate set, each load transfer strategy and the corresponding fault recovery result report in the evaluation model training set are input into the large language model; Use the large language model to determine whether the load transfer strategy meets the evaluation criteria, and output a binary result of compliance or non-compliance; The binarization results of all load transfer strategies under all evaluation criteria are summarized to form an evaluation model assertion set.

[0017] According to a load transfer strategy evaluation device for a power distribution system provided by the present invention, the screening unit is specifically used for: Using statistical devices, the coverage and false rejection rate of the evaluation model assertion set and the expert assertion set are calculated; The consistency index between the assertion set of the evaluation model and the assertion set of the expert is obtained based on the coverage rate and the false rejection rate; The candidate sets of evaluation criteria are sorted according to the consistency index, and the evaluation criterion with the highest consistency index is selected as the optimal evaluation criterion.

[0018] According to a distribution system load transfer strategy evaluation device for expert preference alignment provided by the present invention, the expression of the consistency index is:

[0019]

[0020] In the formula, Represents a set of assertions; for F The coverage rate, for F The false rejection rate, for F Consistency indicators; is the indicator function; if the load transfer strategy , the expert believes that it does not meet the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it does not meet the evaluation criteria provided by the experts, then =1; if the load transfer strategy , the expert believes that it meets the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it meets the evaluation criteria provided by the experts, then =1.

[0021] In a third aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for evaluating load transfer strategies of distribution systems for expert preference alignment according to the first aspect is implemented.

[0022] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the distribution system load transfer strategy evaluation method for expert preference alignment according to the first aspect.

[0023] The present invention provides a distribution system load transfer strategy evaluation method, device, equipment and medium for expert preference alignment, which realizes the distribution system load transfer strategy evaluation method for expert preference alignment, effectively integrates the experience and knowledge of experts into the strategy evaluation process, overcomes the subjectivity and low efficiency of traditional evaluation methods, realizes the automation and intelligence of the evaluation process, improves the accuracy, objectivity and reliability of strategy evaluation, and provides strong support for the safe and stable operation of the distribution system.

[0024] Compared with the prior art, the present invention has the following advantages: 1. Improve evaluation efficiency: By combining the powerful semantic understanding ability of the large language model and expert knowledge, the automatic evaluation of load transfer strategies is realized, which reduces human intervention and improves the efficiency of evaluation.

[0025] 2. Enhanced evaluation accuracy: By screening evaluation criteria that are highly consistent with expert assertions, the evaluation results are ensured to be consistent with expert preferences, thereby improving the accuracy and reliability of the evaluation.

[0026] 3. Quantify expert knowledge: Convert the experience and knowledge of experts into quantifiable evaluation criteria, which solves the problem of difficulty in quantifying expert experience in traditional methods and helps to inherit and share knowledge.

[0027] 4. Improve model generalization ability: By continuously optimizing the evaluation model, collecting more fault recovery cases, and expanding the evaluation model training set, the applicability and generalization ability of the model under different fault scenarios and load transfer strategies are improved.

[0028] 5. Promote human-computer collaboration: The present invention realizes a human-computer collaborative strategy evaluation method, which not only gives full play to the high efficiency of artificial intelligence, but also retains the professional judgment of experts, ensuring the scientificity and rationality of the evaluation results.

[0029] 6. Strong applicability: The present invention is not only applicable to the evaluation of load transfer strategies of distribution systems, but can also be extended to the fields of distribution network planning scheme evaluation, distributed power source access strategy evaluation, load management and demand response strategy evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0031] In the attached picture: Figure 1 A flow chart of a method for evaluating a load transfer strategy of a power distribution system based on expert preference alignment according to the present invention; Figure 2 It is a structural block diagram of the distribution system load transfer strategy evaluation device for expert preference alignment of the present invention; Figure 3 It is a schematic diagram of the structure of the electronic device of the present invention. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Some embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. The following embodiments and features in the embodiments may be combined with each other without conflict. If there is no necessary order for the steps described in the embodiments, the order is only an example and should not be regarded as a limitation. A person of ordinary skill in the art may adjust the order of the steps without destroying the logic.

[0034] Example 1 See also Figure 1 This embodiment provides a method for evaluating a load transfer strategy of a distribution system based on expert preference alignment, which can be based on expert preference alignment and includes the following steps: S101, interacting the load transfer strategy to be evaluated with the simulation environment, generating a fault recovery result report, and jointly constructing an evaluation model training set with the corresponding load transfer strategy; Specifically, the load transfer strategy is used for distribution system fault recovery. However, before use, it needs to be evaluated to predict and evaluate the effect of the load transfer strategy. S101 specifically includes: constructing various possible fault scenarios in a simulation environment, simulating the operation of the distribution system under different fault scenarios, loading and executing the load transfer strategy to be evaluated, and simulating the entire process of distribution system fault recovery; obtaining the operation data, fault recovery process and performance evaluation data of the distribution system to form a fault recovery result report; the fault recovery result report and the corresponding load transfer strategy are jointly constructed into an evaluation model training set. Each sample in the evaluation model training set contains the load transfer strategy itself and the fault recovery result report, which can provide complete data support for the subsequent evaluation process.

[0035] The operation data includes key parameters such as voltage, current, power, frequency, and status information such as switch action and protection action. The performance evaluation data includes key indicators such as fault isolation time, total power supply restoration time, restored load capacity, unrestored load capacity, distribution system loss changes, node voltage deviation, and equipment load rate.

[0036] Total power supply restoration time: the time from the occurrence of a fault to the restoration of power supply to all restorable loads.

[0037] Restored load capacity: The total load capacity to which power supply is successfully restored.

[0038] Unrestored load capacity: List and capacity of loads that cannot have their power restored.

[0039] Changes in distribution system losses: Comparison of distribution system losses before and after the implementation of the load transfer strategy.

[0040] Node voltage deviation: whether the load node voltage is within the allowable range.

[0041] Equipment load rate: whether the power transfer causes line or equipment overload.

[0042] S102. Perform semantic understanding on the text of the regulatory procedures through a preset large language model, generate multiple evaluation criteria in text form, and form a candidate set of evaluation criteria together with the evaluation criteria in text form provided by experts.

[0043] Specifically, the preset large language model may be a pre-trained large language model. The control regulation text includes the operation regulations, fault handling procedures, safety specifications, technical standards, etc. of the power distribution system. Before S102, the control regulation text may also be pre-processed to eliminate noise and redundant information, such as irrelevant comments, format information, etc., and extract valid text content.

[0044] Specifically, in S102, the semantic analysis technology of the large language model is used to extract key judgment criteria related to the evaluation of the load transfer strategy for fault recovery, and generate multiple text-based evaluation judgment criteria for the load transfer strategy for fault recovery (hereinafter referred to as evaluation criteria). These evaluation criteria involve aspects such as safety, reliability, and economy.

[0045] In this embodiment, the evaluation criteria for generating a large language model include: Safety standards: whether the safety range of operating parameters such as current and voltage is met, and whether equipment overload or overheating is avoided.

[0046] Reliability standards: whether the power supply restoration time is minimized and whether important loads are given priority in restoring power.

[0047] Economic standard: whether the network loss of the distribution system is reduced and whether the operating cost is reduced.

[0048] Specifically, in S102, based on their own practical experience and professional knowledge, experts summarize and refine a series of strategy evaluation criteria, such as restrictions on certain key indicators, standard requirements for operation steps, etc., and provide textual evaluation criteria. The evaluation criteria generated by the large language model are combined with the evaluation criteria provided by the experts to form an evaluation criteria candidate set, which contains a variety of possible evaluation criteria and provides a basis for subsequent screening.

[0049] S103. Based on the evaluation criterion candidate set, the evaluation model training set is binarized and evaluated by a large language model to obtain an evaluation model assertion set.

[0050] Specifically, S103 includes: S1031. For each evaluation criterion in the evaluation criterion candidate set, each load transfer strategy and the corresponding fault recovery result report in the evaluation model training set are input into the large language model.

[0051] S1032. Use a large language model to determine whether the load transfer strategy meets the evaluation criteria, and output a binary result of whether it meets or does not meet the criteria.

[0052] S1033: Summarize the binarization results of all load transfer strategies under all evaluation criteria to form an evaluation model assertion set. The evaluation model assertion set records the evaluation results of each load transfer strategy under each evaluation criterion.

[0053] S104. Obtain an expert assertion set formed after the experts perform manual binary evaluation on the evaluation model training set.

[0054] Specifically, experts make comprehensive judgments on the rationality, safety, and effectiveness of the load transfer strategy based on their professional knowledge and experience. Experts review each load transfer strategy and the corresponding fault recovery result report in the evaluation model training set, and give a "compliant" or "non-compliant" evaluation to each load transfer strategy to form an expert assertion set. In order to ensure the objectivity and authority of the evaluation results, multiple experts can evaluate the same load transfer strategy and take the majority opinion or weighted average of their evaluation results.

[0055] S105. Based on the consistency measurement between the expert assertion set and the evaluation model assertion set, the optimal evaluation criterion is selected from the evaluation criterion candidate set.

[0056] Specifically, S105 includes: S1051. Use statistical methods to calculate the coverage and false rejection rate of the evaluation model assertion set and the expert assertion set.

[0057] S1052. Obtain consistency indicators between the evaluation model assertion set and the expert assertion set based on the coverage rate and the false rejection rate.

[0058] Among them, the statistical method is used to calculate the relevant expressions of coverage, false rejection rate (FFR), and alignment indicators as follows:

[0059]

[0060] In the formula, Represents an assertion. y is a binary vector, where {0,1} indicates that experts believe that the load transfer strategy Whether it meets the evaluation criteria provided by the experts (0 means not meeting the criteria, 1 means meeting the criteria). is a set of assertions (including j assertion). is the indicator function; if the load transfer strategy , the expert believes that it does not meet the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it does not meet the evaluation criteria provided by the experts, then =1; if the load transfer strategy , the expert believes that it meets the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it meets the evaluation criteria provided by the experts, then =1.

[0061] S1053. Sort the candidate set of evaluation criteria according to the consistency index, and select the evaluation criteria with the highest consistency index as the optimal evaluation criteria, that is, select the evaluation criteria that are most consistent with the expert evaluation results and determine them as the optimal evaluation criteria.

[0062] In this embodiment, a threshold of the consistency index can be set, and only evaluation criteria with consistency index values ​​higher than the threshold are selected. When multiple evaluation criteria meet the consistency index requirements, the interpretability, operability, and practical application requirements of the evaluation criteria can be comprehensively considered to ultimately determine the optimal evaluation criteria.

[0063] S106. Evaluate the load transfer strategy to be evaluated using the optimal evaluation standard.

[0064] Specifically, the load transfer strategy is automatically evaluated using the optimal evaluation criteria. The load transfer strategy to be evaluated and the fault recovery result report are input into the large language model. According to the optimal evaluation criteria, it is judged whether the load transfer strategy meets the expert preference and the evaluation result is output. The evaluation results can also be fed back to decision makers to provide reference for the fault recovery decision of the distribution system and improve the efficiency and accuracy of the load transfer strategy evaluation.

[0065] Specifically, as the operating environment of the distribution system changes and the text of the control regulations is updated, new evaluation criteria may emerge, and the candidate set of evaluation criteria needs to be updated in a timely manner. In addition, the consistency index is recalculated in combination with the latest expert assertion set to ensure that the evaluation results are continuously aligned with the expert preferences. In addition, more distribution system fault recovery cases are collected, the training set of the evaluation model is expanded, and the generalization ability of the evaluation method is improved so that it can adapt to a wider range of fault scenarios and load transfer strategy types.

[0066] Example 2 like Figure 2 As shown, this embodiment provides a distribution system load transfer strategy evaluation device for expert preference alignment, including: A simulation interaction unit 201 is used to interact the load transfer strategy to be evaluated with the simulation environment, generate a fault recovery result report, and construct an evaluation model training set together with the corresponding load transfer strategy; An evaluation criteria generating unit 202 is used to perform semantic understanding on the text of the regulation procedure through a preset large language model, generate a plurality of evaluation criteria in text form, and form an evaluation criteria candidate set together with the evaluation criteria in text form provided by experts; A first acquisition unit 203 is used to perform a binary evaluation on the evaluation model training set through a large language model based on the evaluation standard candidate set to obtain an evaluation model assertion set; The second acquisition unit 204 is used to acquire an expert assertion set formed after the expert performs manual binary evaluation on the evaluation model training set; A screening unit 205, configured to screen out an optimal evaluation criterion from the evaluation criterion candidate set based on a consistency measure between the expert assertion set and the evaluation model assertion set; The evaluation unit 206 is used to evaluate the load transfer strategy to be evaluated by using an optimal evaluation standard.

[0067] Specifically, the simulation interaction unit 201 is specifically used for: Construct fault scenarios in the simulation environment, load and execute the load transfer strategy to be evaluated; obtain the operating data, fault recovery process and performance evaluation data of the distribution system to form a fault recovery result report; and construct the evaluation model training set together with the fault recovery result report and the corresponding load transfer strategy.

[0068] The text of the control regulations includes the operating procedures, fault handling procedures, safety specifications, and technical standards of the distribution system.

[0069] The first acquisition unit 203 is specifically used for: For each evaluation criterion in the evaluation criterion candidate set, each load transfer strategy and the corresponding fault recovery result report in the evaluation model training set are input into the large language model; Use the large language model to determine whether the load transfer strategy meets the evaluation criteria, and output a binary result of compliance or non-compliance; The binarization results of all load transfer strategies under all evaluation criteria are summarized to form an evaluation model assertion set.

[0070] The screening unit 205 is specifically used for: Using statistical devices, the coverage and false rejection rate of the evaluation model assertion set and the expert assertion set are calculated; The consistency index between the assertion set of the evaluation model and the assertion set of the expert is obtained based on the coverage rate and the false rejection rate; The candidate sets of evaluation criteria are sorted according to the consistency index, and the evaluation criterion with the highest consistency index is selected as the optimal evaluation criterion.

[0071] The expression of consistency index is:

[0072]

[0073] In the formula, Represents a set of assertions; for F The coverage rate, for F The false rejection rate, for F Consistency indicators; is the indicator function; if the load transfer strategy , the expert believes that it does not meet the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it does not meet the evaluation criteria provided by the experts, then =1; if the load transfer strategy , the expert believes that it meets the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it meets the evaluation criteria provided by the experts, then =1.

[0074] Example 3 Figure 3 An example of a structural diagram of an electronic device is shown in FIG. Figure 3As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830 and a communication bus 840, wherein the processor 810, the communication interface 820 and the memory 830 communicate with each other through the communication bus 840. The memory 830 stores a computer program that can be run on the processor 810, and when the processor 810 executes the computer program, the distribution system load transfer strategy evaluation method for expert preference alignment of embodiment 1 is implemented.

[0075] Example 4 This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for evaluating a load transfer strategy of a power distribution system for expert preference alignment of Embodiment 1 is implemented.

[0076] It should be noted that the computer-readable storage medium of the present embodiment may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0077] In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus, or device. In this embodiment, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium may send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The computer program contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0078] The computer readable storage medium may be written in one or more programming languages ​​or a combination thereof to execute the computer program of the present embodiment, and the programming language includes an object-oriented programming language, such as Java, Python, C++, and a conventional procedural programming language, such as C or a similar programming language. The program may be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect through the Internet).

[0079] In summary, the method, device, equipment and medium for evaluating the load transfer strategy of a distribution system for expert preference alignment provided by the present invention adopts a large model evaluation standard screening design, which effectively provides a new solution for the evaluation of large language models, and overcomes the limitations of full automation in the past in understanding complex expert needs and evaluation standards. Ultimately, through manual evaluation by experts, the evaluation standards can be continuously improved and the accuracy of the evaluation can be improved. With the participation of experts, fewer and more accurate assertion sets can be generated, showing a higher evaluation consistency, thereby ensuring the expert preference calibration of the load transfer strategy evaluation to a certain extent, significantly improving the pertinence and practicality of the strategy evaluation, and providing strong support for the optimal scheduling of the distribution system.

[0080] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the embodiments disclosed herein. The present invention is intended to cover any variations, uses or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art that are not disclosed by the present invention. It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for evaluating load transfer strategies in distribution systems based on expert preference alignment, characterized in that: include: The load transfer strategy to be evaluated interacts with the simulation environment to generate a fault recovery result report, and together with the corresponding load transfer strategy, constructs an evaluation model training set; The semantic understanding of the regulatory text is performed through a preset large language model to generate multiple textual evaluation criteria, which together with the textual evaluation criteria provided by experts form a candidate set of evaluation criteria; Based on the evaluation criterion candidate set, the evaluation model training set is binarized and evaluated by the large language model to obtain an evaluation model assertion set; Obtaining an expert assertion set formed by an expert performing manual binary evaluation on the evaluation model training set; Based on the consistency measurement between the expert assertion set and the assessment model assertion set, selecting the optimal evaluation criterion from the evaluation criterion candidate set; The load transfer strategy to be evaluated is evaluated by using the optimal evaluation criteria.

2. The method for evaluating load transfer strategies of distribution systems based on expert preference alignment according to claim 1 is characterized in that: The load transfer strategy to be evaluated interacts with the simulation environment, generates a fault recovery result report, and constructs an evaluation model training set together with the corresponding load transfer strategy, including: Construct a fault scenario in a simulation environment, load and execute the load transfer strategy to be evaluated; obtain the operating data, fault recovery process and performance evaluation data of the distribution system to form the fault recovery result report; and construct the evaluation model training set together with the fault recovery result report and the corresponding load transfer strategy.

3. The method for evaluating load transfer strategies of distribution systems based on expert preference alignment according to claim 1 is characterized in that: The control regulations text includes the operating procedures, fault handling procedures, safety specifications, and technical standards of the distribution system.

4. The method for evaluating load transfer strategies of distribution systems based on expert preference alignment according to claim 1 is characterized in that: The step of performing binary evaluation on the evaluation model training set based on the evaluation standard candidate set by using the large language model to obtain an evaluation model assertion set includes: For each evaluation criterion in the candidate set of evaluation criteria, inputting each load transfer strategy and the corresponding fault recovery result report in the evaluation model training set into the large language model; Using the large language model to determine whether the load transfer strategy meets the evaluation criteria, and outputting a binary result of whether it meets or does not meet the criteria; The binarization results of all load transfer strategies under all evaluation criteria are summarized to form the evaluation model assertion set.

5. The method for evaluating load transfer strategies of distribution systems based on expert preference alignment according to claim 1 is characterized in that: The method of screening out the optimal evaluation criterion from the evaluation criterion candidate set based on the consistency measurement between the expert assertion set and the evaluation model assertion set includes: Using statistical methods, calculate the coverage and false rejection rate of the assessment model assertion set and the expert assertion set; Obtaining a consistency index between the assessment model assertion set and the expert assertion set according to the coverage rate and the false rejection rate; The candidate set of evaluation criteria is sorted according to the consistency index, and the evaluation criterion with the highest consistency index is selected as the optimal evaluation criterion.

6. The method for evaluating load transfer strategies of distribution systems based on expert preference alignment according to claim 5 is characterized in that: The expression of the consistency index is: In the formula, Represents a set of assertions; for F The coverage rate, for F The false rejection rate, for F Consistency indicators; is the indicator function; if the load transfer strategy , the expert believes that it does not meet the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it does not meet the evaluation criteria provided by the experts, then =1; if the load transfer strategy , the expert believes that it meets the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it meets the evaluation criteria provided by the experts, then =1.

7. A distribution system load transfer strategy evaluation device for expert preference alignment, characterized in that: include: A simulation interaction unit is used to interact the load transfer strategy to be evaluated with the simulation environment, generate a fault recovery result report, and jointly construct an evaluation model training set with the corresponding load transfer strategy; An evaluation criteria generation unit, which is used to perform semantic understanding of the regulatory procedure text through a preset large language model, generate multiple evaluation criteria in text form, and form an evaluation criteria candidate set together with the evaluation criteria in text form provided by experts; A first acquisition unit is used to perform a binary evaluation on the evaluation model training set through the large language model based on the evaluation criterion candidate set to obtain an evaluation model assertion set; A second acquisition unit is used to acquire an expert assertion set formed after an expert performs manual binary evaluation on the evaluation model training set; A screening unit, configured to screen out an optimal evaluation criterion from the evaluation criterion candidate set based on a consistency measure between the expert assertion set and the evaluation model assertion set; An evaluation unit is used to evaluate the load transfer strategy to be evaluated according to the optimal evaluation standard.

8. The distribution system load transfer strategy evaluation device for expert preference alignment according to claim 7, characterized in that: The simulation interaction unit is specifically used for: Construct a fault scenario in a simulation environment, load and execute the load transfer strategy to be evaluated; obtain operation data, fault recovery process and performance evaluation data of the power distribution system, and form the fault recovery result report; The fault recovery result report and the corresponding load transfer strategy are used together to construct the evaluation model training set.

9. The distribution system load transfer strategy evaluation device for expert preference alignment according to claim 7, characterized in that: The control regulations text includes the operating procedures, fault handling procedures, safety specifications, and technical standards of the distribution system.

10. The distribution system load transfer strategy evaluation device for expert preference alignment according to claim 7, characterized in that: The first acquisition unit is specifically used for: For each evaluation criterion in the candidate set of evaluation criteria, inputting each load transfer strategy and the corresponding fault recovery result report in the evaluation model training set into the large language model; Using the large language model to determine whether the load transfer strategy meets the evaluation criteria, and outputting a binary result of whether it meets or does not meet the criteria; The binarization results of all load transfer strategies under all evaluation criteria are summarized to form the evaluation model assertion set.

11. The distribution system load transfer strategy evaluation device for expert preference alignment according to claim 7, characterized in that: The screening unit is specifically used for: Using a statistical device, calculating the coverage and false rejection rate of the assessment model assertion set and the expert assertion set; Obtaining a consistency index between the assessment model assertion set and the expert assertion set according to the coverage rate and the false rejection rate; The candidate set of evaluation criteria is sorted according to the consistency index, and the evaluation criterion with the highest consistency index is selected as the optimal evaluation criterion.

12. The distribution system load transfer strategy evaluation device for expert preference alignment according to claim 11, characterized in that: The expression of the consistency index is: In the formula, Represents a set of assertions; for F The coverage rate, for F The false rejection rate, for F Consistency indicators; is the indicator function; if the load transfer strategy , the expert believes that it does not meet the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it does not meet the evaluation criteria provided by the experts, then =1; if the load transfer strategy , the expert believes that it meets the evaluation criteria provided by the expert, and the large language model believes that it does not meet one of the evaluation criteria provided by the large language model, then =1; if the load transfer strategy , the experts believe that it meets the evaluation criteria provided by the experts, then =1.

13. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for evaluating the load transfer strategy of a distribution system based on expert preference alignment as described in any one of claims 1 to 6 is implemented.

14. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for evaluating a load transfer strategy of a distribution system based on expert preference alignment as described in any one of claims 1 to 6 is implemented.

Images