CN-121998450-A - High-energy-consumption industrial user load regulation value evaluation method, system and medium

CN121998450ACN 121998450 ACN121998450 ACN 121998450ACN-121998450-A

Abstract

A high-energy-consumption industrial user load regulation and control value evaluation method, system and medium comprise the steps of generating a candidate scheme through a preset regulation and control strategy based on an operation plan and constraint conditions of a high-energy-consumption industrial user, inputting the candidate scheme into a pre-built multi-objective similarity evaluation model to obtain a correlation coefficient of the candidate scheme relative to a positive/negative reference sequence, determining deviation of the candidate scheme relative to the positive/negative reference sequence based on the correlation coefficient, and carrying out risk value evaluation based on the deviation of the candidate scheme and the positive/negative reference sequence and a pre-built risk sensitive perception model. The multi-objective similarity evaluation model can integrate effective information through the association coefficient under the scene of incomplete information and uncertain system, avoid evaluation failure caused by data loss of the traditional model, and accurately describe subjective perception of a decision maker on regulation and control income and regulation and control loss through a perception value function and a decision weight function of a risk sensitive perception model.

Inventors

ZHENG BOWEN
XU ZISHANG
YANG MINGYU
LI YUYING
GUO ZIXUAN
LIU CHANG
PAN MINGMING
ZHUANG ZHONG
DUAN MEIMEI
FANG KAIJIE
HUANG YIXUAN
LI YONGJUN
YUAN JINDOU

Assignees

中国电力科学研究院有限公司
国网江苏省电力有限公司
国家电网有限公司

Dates

Publication Date: 20260508
Application Date: 20251226

Claims (16)

1. The method for evaluating the load regulation value of the high-energy-consumption industrial user is characterized by comprising the following steps of: forming a candidate scheme by a preset regulation strategy generation rule based on an operation plan and constraint conditions of a high-energy-consumption industrial user, wherein the constraint conditions comprise equipment start-stop constraint, production interruptibility and load transfer capacity constraint; Inputting the candidate scheme into a pre-constructed multi-objective similarity evaluation model to obtain a correlation coefficient of the candidate scheme relative to the positive/negative reference sequence, and determining the deviation amount of the candidate scheme relative to the positive/negative reference sequence based on the correlation coefficient; performing risk value assessment based on the deviation of the candidate scheme and the positive/negative reference sequence and a pre-constructed risk sensitive perception model; the constructed multi-target similarity evaluation model is constructed by determining an ideal reference vector according to target attributes and industry physical rules in a high-energy-consumption industry user load regulation scene and combining a similarity measurement method.
2. The method of claim 1, wherein the constructing of the multi-objective similarity assessment model comprises: determining positive/negative reference coefficients according to various calculation formulas according to index types in a high-energy-consumption industrial user load regulation scene, and constructing positive/negative reference sequences according to time sequence based on the positive/negative reference coefficients; And constructing a multi-objective similarity evaluation model by taking the positive reference sequence and the negative reference sequence as the basis of decision and combining a similarity measurement method.
3. The method of claim 1, wherein inputting the candidate solution into a pre-constructed multi-objective similarity assessment model to obtain a correlation coefficient of the candidate solution with respect to the positive/negative reference sequence, and determining an offset of the candidate solution with respect to the positive/negative reference sequence based on the correlation coefficient, comprises: Carrying out dimension-by-dimension comparison on the numerical value of each candidate scheme in the multi-dimensional space and the positive/negative reference sequence in the multi-objective similarity evaluation model by adopting a similarity measurement method to obtain the correlation coefficient of each candidate scheme relative to the positive/negative reference sequence in the multi-dimensional space, and converting the correlation coefficient of each candidate scheme relative to the positive/negative reference sequence in the multi-dimensional space into a standardized similarity coefficient; The normalized similarity coefficient is taken as the offset of the candidate from the positive/negative reference sequence.
4. The method of claim 3, wherein inputting the candidate solution into a pre-constructed multi-objective similarity assessment model to obtain a correlation coefficient of the candidate solution with the positive/negative reference sequence, and determining an offset of the candidate solution relative to the positive/negative reference sequence based on the correlation coefficient, further comprises: and integrating the standardized similarity coefficients of each candidate scheme in each dimensional space to obtain the comprehensive association coefficient of the candidate scheme relative to the ideal reference vector.
5. A method according to claim 3, wherein the normalized similarity coefficient is calculated as: Wherein, the For the correlation coefficient of scheme i under index k, Is the value of scheme i at index k, Is the value of the ideal reference vector at index k, Is the resolution factor.
6. The method of claim 4, wherein the integrated correlation coefficient of the candidate with respect to the ideal reference vector is calculated as: Where n is the total number of indicators, Is the correlation coefficient of scheme i under index k, In order to integrate the correlation coefficient(s), The index number is i, and the enterprise number is i.
7. The method of claim 1, wherein the risk value assessment based on the deviation of the candidate solution from the positive/negative reference sequence in combination with a pre-constructed risk sensitive perceptual model comprises: Substituting the deviation of the candidate scheme and the positive/negative reference sequence into a positive/negative risk cost function in a pre-constructed risk sensitive perception model to obtain the positive/negative risk value of the index relative to the high-energy-consumption industrial user under the overhaul regulation value; Determining decision weights facing to benefits/losses based on weights of indexes under overhaul regulation values and subjective weight functions of a pre-constructed risk sensitive perception model; Determining overhaul regulation values of candidate schemes based on the positive/negative risk value in combination with decision weights in the face of profit/loss; and carrying out risk value assessment based on the overhaul regulation value of the candidate scheme.
8. The method of claim 7, wherein the positive/negative risk cost function is represented by the formula: in the formula, For a positive risk cost function of the index j relative to the high energy industrial user i under the maintenance and control value, For the maintenance of a negative risk cost function of the index j relative to the high energy industrial user i under the regulation value, For overhauling the negative reference coefficient of the j index of the high energy consumption industrial user i under the regulation value, For overhauling the positive reference coefficient of the j index of the high energy consumption industrial user i under the regulation value, And The risk preference coefficient and the risk avoidance coefficient are respectively provided, i is the serial number of the high-energy-consumption industrial user, j is the serial number of the index, The sensitivity coefficient of decision maker to gain and loss.
9. The method of claim 7, wherein the subjective weight function is represented by the formula: in the formula, As the decision weight in the case of benefit, In order to make a decision weight at the time of loss, The weight of the j index under the overhaul regulation value; And Risk attitude coefficients for decision makers facing returns and losses respectively, 、 The sensitivity of the decision maker to the residual weight under the situation of benefit and loss is respectively described.
10. The method of claim 7, wherein the service control value is calculated as follows: in the formula, Is the overhaul regulation value of the high-energy-consumption industrial user i, For a positive risk cost function of the index j relative to the high energy industrial user i under the maintenance and control value, For the maintenance of a negative risk cost function of the index j relative to the high energy industrial user i under the regulation value, For decision weights when faced with benefits, And m is the index number for decision weights when loss is faced.
11. A high energy industrial consumer load regulation value assessment system, comprising: the scheme generation module is used for generating rules to form candidate schemes through a preset regulation and control strategy based on an operation plan and constraint conditions of high-energy-consumption industrial users, wherein the constraint conditions comprise equipment start-stop constraint, production interruptibility and load transfer capacity constraint; The similarity evaluation module is used for inputting the candidate schemes into a pre-built multi-objective similarity evaluation model, obtaining association coefficients of the candidate schemes relative to the positive/negative reference sequences, and determining the deviation amount of the candidate schemes relative to the positive/negative reference sequences based on the association coefficients; The value evaluation module is used for evaluating the risk value based on the deviation of the candidate scheme and the positive/negative reference sequence and combining a pre-constructed risk sensitive perception model; the constructed multi-target similarity evaluation model is constructed by determining an ideal reference vector according to target attributes and industry physical rules in a high-energy-consumption industry user load regulation scene and combining a similarity measurement method.
12. The system of claim 11, wherein the similarity evaluation module comprises: the computing sub-module is used for carrying out dimension-by-dimension comparison on the numerical value of each candidate scheme in the multi-dimensional space and the positive/negative reference sequence in the multi-objective similarity evaluation model by adopting a similarity measurement method to obtain the association coefficient of each candidate scheme relative to the positive/negative reference sequence in the multi-dimensional space; And the normalization sub-module is used for converting the association coefficient of each candidate scheme relative to the positive/negative reference sequence in a multidimensional space into a normalized similarity coefficient, and taking the normalized similarity coefficient as the deviation of the candidate scheme from the positive/negative reference sequence.
13. The system of claim 11, wherein the value assessment module is specifically configured to: Substituting the deviation of the candidate scheme and the positive/negative reference sequence into a positive/negative risk cost function in a pre-constructed risk sensitive perception model to obtain the positive/negative risk value of the index relative to the high-energy-consumption industrial user under the overhaul regulation value; Determining decision weights facing to benefits/losses based on weights of indexes under overhaul regulation values and subjective weight functions of a pre-constructed risk sensitive perception model; Determining overhaul regulation values of candidate schemes based on the positive/negative risk value in combination with decision weights in the face of profit/loss; and carrying out risk value assessment based on the overhaul regulation value of the candidate scheme.
14. The system of claim 11, wherein the subjective weight function is represented by the formula: in the formula, As the decision weight in the case of benefit, In order to make a decision weight at the time of loss, The weight of the j index under the overhaul regulation value; And Risk attitude coefficients for decision makers facing returns and losses respectively, 、 The sensitivity of the decision maker to the residual weight under the situation of benefit and loss is respectively described.
15. The electronic equipment is characterized by comprising at least one processor and a memory, wherein the memory and the processor are connected through a bus; The memory is used for storing one or more programs; a high energy industrial user load regulation value assessment method according to any one of claims 1 to 10, when said one or more programs are executed by said at least one processor.
16. A readable storage medium having stored thereon an execution program which, when executed, implements a high energy industrial user load regulation value evaluation method according to any one of claims 1 to 10.

Description

High-energy-consumption industrial user load regulation value evaluation method, system and medium Technical Field The invention relates to the field of control of power systems, in particular to a high-energy-consumption industrial user load regulation value evaluation method, a system and a medium. Background Along with the transformation of the power system to a novel power system, high-energy-consumption industrial users (such as industries of steel, chemical industry, nonferrous metals and the like) are taken as a core load main body of the power system, and the load controllability of the power system has important significance for peak load regulation, new energy consumption and stable operation of the power grid. At present, the load regulation and control of the power system on high-energy-consumption industrial users is updated from 'passive peak clipping' to 'active optimization', and the optimal regulation and control strategy is determined by accurately quantifying the values of different regulation and control means (maintenance, rotation, time-shifting and peak avoidance). The current related key technology focuses on execution of load regulation and control means and single-dimensional value evaluation, wherein on one hand, the existing regulation and control means cover the scenes of overhaul (load adjustment during equipment maintenance), rotation (load transfer during periodical production stoppage), time-shifting (production period translation), peak avoidance (load reduction during electricity consumption peak period) and the like, but lack quantitative basis for the value of different means, on the other hand, the traditional load value evaluation method mainly adopts single-objective evaluation (such as only considering load reduction amount or economic cost), or relies on complete and determined operation data, and is difficult to adapt to the actual scenes of fragmentation of production data and uncertainty of equipment states (such as load fluctuation caused by sudden faults) of high-energy industrial users. The prior art has the obvious defects that: The traditional evaluation model depends on complete index data, and in the scene of incomplete information (such as partial production data loss) of high-energy-consumption industrial users and uncertain system state (such as load affected by fluctuation of raw material supply), the evaluation result has larger deviation and cannot support reliable decision; Risk perception loss, namely ignoring risk preference of a decision maker in load regulation (for example, the evasive tendency of 'regulation leading to production loss' is stronger than the pursuit of 'regulation leading to subsidy gain'), and the evaluation result is disjointed with the actual decision requirement, so that the execution resistance of a regulation scheme is easy to lead to great; The reference system is single, a single standard sequence (such as 'optimal load reduction') is adopted as an evaluation reference, positive and negative effects of indexes (such as 'load reduction' not only reduces electricity consumption cost, but also possibly reduces productivity) are not considered, the problem that the dimensions of different indexes (such as economic cost, production efficiency and power grid contribution) are not uniform is not solved, and the evaluation precision is low; the quantization logic is incomplete, and the value of different means cannot be compared transversely due to the lack of a unified quantization frame for different control means such as overhaul, rotation and the like, so that the distribution of control resources is unreasonable. Disclosure of Invention In order to solve the problems of poor information adaptability, risk perception deficiency, single reference system, incomplete quantization logic and the like in the prior art, the invention provides a high-energy-consumption industrial user load regulation and control value evaluation method, which comprises the following steps: forming a candidate scheme by a preset regulation strategy generation rule based on an operation plan and constraint conditions of a high-energy-consumption industrial user, wherein the constraint conditions comprise equipment start-stop constraint, production interruptibility and load transfer capacity constraint; Inputting the candidate scheme into a pre-constructed multi-objective similarity evaluation model to obtain a correlation coefficient of the candidate scheme relative to the positive/negative reference sequence, and determining the deviation amount of the candidate scheme relative to the positive/negative reference sequence based on the correlation coefficient; performing risk value assessment based on the deviation of the candidate scheme and the positive/negative reference sequence and a pre-constructed risk sensitive perception model; the constructed multi-target similarity evaluation model is constructed by determining an ideal reference vector acco