Power dispatching method and system based on energy computing power interconnection

Through the method based on energy computing power interconnection, load prediction and power scheduling decision mining are used to use predicted feature time zones and environmental data, combined with iterative optimization and in-depth analysis, the problem of low efficiency of existing power scheduling methods is solved, and the power system is improved Efficiency and reliability.

CN119005606AInactive Publication Date: 2024-11-22STATE GRID JIANGSU ELECTRIC POWER CO LTD NANTONG POWER SUPPLY BRANCH +3

Patent Information

Application Number
CN202411075247.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-07
Publication Date
2024-11-22
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Due to the lack of rigorous and completeness of the existing power scheduling method, the power scheduling efficiency is low and the efficiency and reliability of the power system cannot be improved.

Method used

Using the method based on energy computing power interconnection, load prediction is predicted by reading predicted feature time zones and predicted feature environment data, power scheduling decision space is mined, and pre-built energy computing power interconnection model is used for iterative optimization and in-depth optimization analysis is carried out. Obtain a power dispatch plan.

Benefits of technology

It improves the efficiency and reliability of the power system and solves the problem of low power scheduling efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119005606A_ABST
    Figure CN119005606A_ABST
Patent Text Reader

Abstract

The invention discloses a power dispatching method and system based on energy computing power interconnection, and relates to the technical field of power dispatching, and the method comprises the steps: reading a prediction feature time zone and prediction feature environment data corresponding to the prediction feature time zone; performing load prediction on the target power system to obtain a target load prediction result; performing power dispatching decision mining on the target power system to obtain a power dispatching decision space; based on a predetermined decision quality constraint and a predetermined optimization convergence constraint, performing iterative optimization on the power dispatching decision space according to a pre-constructed energy computing power interconnection model to obtain a power dispatching decision optimization domain; performing deep optimization analysis to obtain a power dispatching scheme; and performing power dispatching on the target power system, thereby solving the problem of low power dispatching efficiency caused by insufficient preciseness and completeness of the power dispatching work in the prior art, and improving the efficiency and reliability of the final power system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of electric power dispatching, and in particular to an electric power dispatching method and system based on energy computing power interconnection. Background Art

[0002] As the scale and complexity of power systems continue to increase, traditional power dispatching methods can no longer meet the requirements of high efficiency, reliability, and sustainable development. In order to solve this problem, energy computing interconnection has emerged. Energy computing interconnection is a new type of energy intelligent technology. Based on advanced computer science and control theory, it interconnects and coordinates various nodes in the power system, such as power plants, transmission lines, and substations, to achieve global optimization and intelligent management.

[0003] The existing power dispatching technology is not rigorous and complete enough, which leads to low power dispatching efficiency, and ultimately makes it impossible to improve the efficiency and reliability of the power system. Summary of the invention

[0004] The present application provides a power dispatching method and system based on energy computing power interconnection, which solves the problem of low power dispatching efficiency due to insufficient rigor and completeness in the prior art, thereby ultimately improving the efficiency and reliability of the power system.

[0005] In view of the above problems, the present application provides a power dispatching method based on energy computing power interconnection.

[0006] In the first aspect, the present application provides a power dispatching method based on energy computing power interconnection, the method comprising: reading a predicted characteristic time zone, and predicted characteristic environmental data corresponding to the predicted characteristic time zone; performing load forecasting on a target power system according to the predicted characteristic time zone and the predicted characteristic environmental data to obtain a target load forecast result; performing power dispatching decision mining on the target power system based on the target load forecast result to obtain a power dispatching decision space; based on predetermined decision quality constraints and predetermined optimization convergence constraints, iteratively optimizing the power dispatching decision space according to a pre-constructed energy computing power interconnection model to obtain a power dispatching decision optimization domain; performing deep optimization analysis based on the power dispatching decision optimization domain to obtain a power dispatching plan; and performing power dispatching on the target power system based on the power dispatching plan.

[0007] In the second aspect, the present application provides an electric power dispatching system based on energy computing power interconnection, and the system includes: a prediction feature time zone module: reading the prediction feature time zone, and the prediction feature environmental data corresponding to the prediction feature time zone; a load prediction result module: performing load prediction on the target power system according to the prediction feature time zone and the prediction feature environmental data, and obtaining a target load prediction result; a scheduling decision space module: performing electric power dispatching decision mining on the target power system based on the target load prediction result, and obtaining an electric power dispatching decision space; an optimization convergence constraint module: based on predetermined decision quality constraints and predetermined optimization convergence constraints, iteratively optimizing the electric power dispatching decision space according to a pre-built energy computing power interconnection model, and obtaining an electric power dispatching decision optimization domain; a deep optimization analysis module: performing deep optimization analysis based on the electric power dispatching decision optimization domain to obtain an electric power dispatching plan; an electric power system dispatching module: performing electric power dispatching on the target electric power system based on the electric power dispatching plan.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0009] The power dispatching method and system based on energy computing power interconnection provided in the embodiment of the present application reads the predicted feature time zone and the predicted feature environmental data corresponding to the predicted feature time zone, and performs load forecasting on the target power system to obtain the target load forecasting result, and then performs power dispatching decision mining on the target power system based on the target load forecasting result to obtain the power dispatching decision space, and based on predetermined decision quality constraints and predetermined optimization convergence constraints, iteratively optimizes the power dispatching decision space according to a pre-built energy computing power interconnection model to obtain the power dispatching decision optimization domain, and finally performs in-depth optimization analysis based on the power dispatching decision optimization domain to obtain the power dispatching plan, and performs power dispatching on the target power system, thereby solving the problem of low power dispatching efficiency due to insufficient rigor and completeness of power dispatching work in the prior art, thereby improving the efficiency and reliability of the final power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A flow chart of a power dispatching method based on energy computing power interconnection is provided for this application;

[0011] Figure 2 A schematic diagram of the structure of an electric power dispatching system based on energy computing power interconnection is provided for this application.

[0012] Explanation of the accompanying drawings: prediction feature time zone module 11, load prediction result module 12, scheduling decision space module 13, optimization convergence constraint module 14, deep optimization analysis module 15, power system scheduling module 16. DETAILED DESCRIPTION

[0013] The present application provides a power dispatching method and system based on energy computing power interconnection, reads the predicted characteristic time zone, and the predicted characteristic environmental data corresponding to the predicted characteristic time zone, and performs load forecasting on the target power system to obtain the target load forecast result, and then performs power dispatching decision mining on the target power system based on the target load forecast result to obtain the power dispatching decision space, and based on the predetermined decision quality constraints and the predetermined optimization convergence constraints, iteratively optimizes the power dispatching decision space according to the pre-built energy computing power interconnection model to obtain the power dispatching decision optimization domain, and finally performs in-depth optimization analysis based on the power dispatching decision optimization domain to obtain the power dispatching plan, and performs power dispatching on the target power system. The problem of low power dispatching efficiency due to insufficient rigor and completeness of power dispatching work in the prior art is solved, so that the efficiency and reliability of the final power system are improved.

[0014] Embodiment 1

[0015] like Figure 1 As shown, the present application provides a power dispatching method and system based on energy computing power interconnection, the method comprising:

[0016] Reading the predicted characteristic time zone and the predicted characteristic environment data corresponding to the predicted characteristic time zone;

[0017] Perform load forecasting on the target power system according to the forecast characteristic time zone and the forecast characteristic environmental data to obtain a target load forecast result;

[0018] Obtaining a load record data set of the target power system;

[0019] Performing principal component analysis on the load record data set to obtain a dimension-reduced load record data set;

[0020] Based on the Markov chain, data integration is performed on the dimension-reduced load record data set to obtain a load forecast chain;

[0021] The prediction characteristic time zone and the prediction characteristic environment data are input into the load prediction chain to obtain the target load prediction result.

[0022] The prediction feature time zone refers to the time range of the predicted load, and the prediction feature environment data is the prediction feature environment corresponding to the prediction feature time zone, including historical load data, weather data, etc. The prediction feature time zone to be used is determined, and then the load forecast of the target power system is carried out. First, it is necessary to obtain the historical load data set of the target power system, access the storage data and traverse the historical load data set to obtain the historical load data set. The historical load data set is a data set that records the load conditions of the power system in the past period of time, usually presented in the form of a time series. Principal component analysis is a commonly used data dimensionality reduction technology. It maps the original data from a high-dimensional space to a low-dimensional space through linear transformation and retains the most important features. PCA is used to reduce the dimension of the load record data set to reduce the dimension of the data and obtain a reduced-dimensional load record data set. The reduced-dimensional load record data set is integrated based on the Markov chain. The Markov chain is a random process model with memory that can describe the transition probability of a system between different states. The Markov chain model is used to integrate the reduced-dimensional load record data set to establish a load forecast chain. Finally, the predicted characteristic time zone and predicted characteristic environmental data are taken as input and combined with the load forecasting chain. The forecast result of the target load is calculated through the Markov chain model, which provides a data basis for the subsequent power dispatching decision mining of the target power system based on the target load forecasting results and obtaining the power dispatching decision space.

[0023] Based on the target load forecast result, power dispatch decision mining is performed on the target power system to obtain a power dispatch decision space;

[0024] Reading the power dispatch decision record of the target power system according to the target load forecast result to obtain a power dispatch decision record set;

[0025] Calibrate the trigger interval of the power dispatch decision indicator based on the power dispatch decision record set to obtain multiple power dispatch decision indicator trigger intervals;

[0026] Traversing the multiple power dispatch decision indicator triggering intervals to perform multiple random value acquisitions to obtain multiple power dispatch decisions;

[0027] The plurality of power dispatch decisions are added to the power dispatch decision space.

[0028] Power dispatch decision mining refers to the mining and analysis of power dispatch decisions based on the target load forecast results, including considering the operation constraints, dispatch objectives and feasible dispatch plans of the power system to determine a reasonable power dispatch decision space. Specifically, according to the target load forecast results, the past power dispatch decision record set, that is, the power dispatch decision record, is read. These records include the power dispatch decisions made in history for similar forecast situations, which can be used as a reference and reference to obtain the power dispatch decision record set. Based on the power dispatch decision record set, the trigger conditions and ranges of different power dispatch decisions are determined by analyzing and judging key indicators, and the trigger interval calibration of the power dispatch decision indicators is completed. And multiple random values ​​are taken in the trigger intervals of multiple power dispatch decision indicators. Multiple dispatch plans are obtained by simulating power dispatch decisions under different circumstances. Multiple power dispatch decision plans are obtained by traversing multiple power dispatch decision indicator trigger intervals and taking multiple random values, and multiple power dispatch decisions are added to the power dispatch decision space. Power dispatch decision mining for the target power system through the target load forecast results can obtain power dispatch decision spaces for various different situations.

[0029] Based on the predetermined decision quality constraint and the predetermined optimization convergence constraint, the power dispatch decision space is iteratively optimized according to the pre-built energy computing power interconnection model to obtain the power dispatch decision optimization domain;

[0030] Performing random extraction according to the power dispatch decision space to obtain a first power dispatch decision;

[0031] Calculating a first decision prediction quality coefficient corresponding to the first power dispatch decision according to the energy computing power interconnection model;

[0032] Determining whether the first decision prediction quality coefficient satisfies the predetermined decision quality constraint;

[0033] If the first decision prediction quality coefficient satisfies the predetermined decision quality constraint, adding the first power dispatch decision to the power dispatch decision optimization domain;

[0034] Based on the predetermined decision quality constraint and the energy computing power interconnection model, the power dispatch decision space is continuously iterated and optimized until the power dispatch decision optimization domain that satisfies the predetermined optimization convergence constraint is generated.

[0035] The predetermined decision quality constraint refers to the constraint of the preset decision prediction quality coefficient range, and the predetermined optimization convergence constraint refers to the constraint of the preset decision number. The energy computing power interconnection model is an iterative optimization calculation model built on a neural network. A power dispatching scheme is randomly selected from the previously constructed power dispatching decision space as the first power dispatching decision, and the energy computing power interconnection model is used to calculate the first power dispatching decision to obtain its corresponding decision prediction quality coefficient. Among them, the decision prediction quality coefficient can evaluate the prediction accuracy and feasibility of the power dispatching decision. According to the predetermined decision quality constraint, it is judged whether the first decision prediction quality coefficient meets the requirements. If it meets the requirements, the first power dispatching decision is added to the power dispatching decision optimization domain as an alternative solution that meets the predetermined decision quality constraint; if it does not meet the requirements, the power dispatching decision is reselected for calculation. According to the predetermined decision quality constraint and the energy computing power interconnection model, the power dispatching decision optimization domain is iteratively optimized. In each iteration, a power dispatching scheme is randomly extracted from the power dispatching decision space, its decision prediction quality coefficient is calculated, and it is judged whether it meets the predetermined decision quality constraint. If it meets the requirements, the power dispatching decision is added to the power dispatching decision optimization domain. Finally, when the power dispatch decision optimization domain meets the predetermined optimization convergence constraint, the optimization process ends. By iteratively optimizing the power dispatch decision space, the constructed power dispatch decision optimization domain can be made more accurate.

[0036] Performing in-depth optimization analysis based on the power dispatch decision optimization domain to obtain a power dispatch plan;

[0037] The target power system is subjected to power dispatch based on the power dispatch plan.

[0038] According to the power dispatch decision optimization domain, a first winning power dispatch decision and a first winning decision prediction quality coefficient corresponding to the first winning power dispatch decision are selected;

[0039] Randomly adjusting each power dispatching decision in the power dispatching decision optimization domain to obtain multiple power dispatching decision neighborhoods;

[0040] Based on the energy computing power interconnection model, calculating multiple neighborhood-decision prediction quality coefficients corresponding to the multiple power dispatch decision neighborhoods;

[0041] Based on the multiple neighborhood-decision prediction quality coefficients and the multiple power dispatch decision neighborhoods, screening the optimal neighborhood-decision prediction quality coefficient and the optimal neighborhood-power dispatch decision corresponding to the optimal neighborhood-decision prediction quality coefficient;

[0042] Determining whether the optimal neighborhood-decision prediction quality coefficient is greater than the first winning decision prediction quality coefficient;

[0043] If the optimal neighborhood-decision prediction quality coefficient is greater than the first winning decision prediction quality coefficient, outputting the optimal neighborhood-power dispatch decision as the power dispatch plan;

[0044] If the optimal neighborhood-decision prediction quality coefficient is less than / equal to the first winning decision prediction quality coefficient, the first winning power dispatching decision is output as the power dispatching plan.

[0045] According to the power dispatch decision optimization domain, a deep optimization analysis is performed. First, a power dispatch decision optimization domain is generated according to the energy computing power interconnection model, and the first winning power dispatch decision and the corresponding first winning decision prediction quality coefficient are screened out. Then, each power dispatch decision in the power dispatch decision optimization domain is randomly adjusted to obtain multiple power dispatch decision neighborhoods. The energy computing power interconnection model is then used to calculate the neighborhood-decision prediction quality coefficients corresponding to these power dispatch decision neighborhoods. After obtaining multiple neighborhood-decision prediction quality coefficients and corresponding power dispatch decision neighborhoods, the optimal neighborhood-decision prediction quality coefficient and the corresponding optimal neighborhood-power dispatch decision are screened out. Then, it is determined whether the optimal neighborhood-decision prediction quality coefficient is greater than the first winning decision prediction quality coefficient. If the optimal neighborhood-decision prediction quality coefficient is greater than the first winning decision prediction quality coefficient, then the optimal neighborhood-power dispatch decision is output as the power dispatch plan. If the optimal neighborhood-decision prediction quality coefficient is less than or equal to the first winning decision prediction quality coefficient, then the first winning power dispatch decision is output as the power dispatch plan, and the power dispatch plan is obtained, and the target power system is dispatched according to the power dispatch plan. Conducting in-depth optimization analysis based on the power dispatch decision optimization domain to find the best power dispatch solution can improve the efficiency and quality of the power system.

[0046] Furthermore, the method further comprises:

[0047] Modeling the target power system based on the digital twin platform to obtain a power simulation model;

[0048] According to the first power dispatching decision, the power simulation model is simulated for power dispatching to obtain a first decision-simulated operating condition data set;

[0049] The energy computing power interconnection model includes an input layer, a simulation scheduling evaluation layer, an evaluation incentive adjustment layer and an output layer;

[0050] Inputting the first decision simulation working condition data set into the simulation scheduling evaluation layer to obtain a first simulation scheduling evaluation result, wherein the simulation scheduling evaluation layer includes multi-dimensional simulation scheduling evaluation indicators, and the multi-dimensional simulation scheduling evaluation indicators include power scheduling safety, power scheduling energy utilization, power scheduling load adaptability and power scheduling stability;

[0051] The first simulation scheduling evaluation result is input into the evaluation incentive adjustment layer to obtain the first decision prediction quality coefficient, wherein the evaluation incentive adjustment layer includes preset evaluation incentive adjustment constraints.

[0052] The digital twin platform is a comprehensive simulation platform based on models and data, which can map the physical, chemical and biological processes of the actual system to the computer in a digital way. The target power system is modeled through the digital twin platform to create a power simulation model that accurately reflects the characteristics and behaviors of the actual system. The first power dispatch decision is applied to the power simulation model to simulate power dispatch. According to the simulated operation of power dispatch, a first decision simulation condition data set is generated to record the values ​​of various variables and the operating status of the system during the simulated power dispatch process. The energy computing power interconnection model includes an input layer, a simulation dispatch evaluation layer, an evaluation incentive adjustment layer and an output layer. Among them, the input layer of the energy computing power interconnection model is used to receive the first power dispatch simulation condition data set as input; the simulation dispatch evaluation layer is to evaluate the quality of the simulation dispatch by applying a series of evaluation indicators. These evaluation indicators may include power dispatch safety, power dispatch energy utilization, power dispatch load adaptability, power dispatch stability and other aspects. Each evaluation indicator will be calculated and analyzed based on the input simulation condition data; in the evaluation incentive adjustment layer, according to the preset evaluation incentive adjustment constraints, the first decision prediction quality coefficient is calculated and obtained. These constraints are set based on actual needs and requirements. They are obtained by assigning a weight to each of the power dispatching safety, power dispatching energy utilization, power dispatching load adaptability and power dispatching stability, and performing weighted calculations to limit and adjust the quality of the simulated dispatching. The evaluation incentive adjustment layer can help determine the quality of the simulated dispatching and provide a measurement indicator; the output layer of the energy computing power interconnection model produces the first decision prediction quality coefficient as the output result, which can be used to measure the quality of the simulated dispatching, thereby supporting more accurate and reasonable power dispatching decisions. The input layer, simulation dispatching evaluation layer, evaluation incentive adjustment layer and output layer are constructed respectively, and the constructed input layer, simulation dispatching evaluation layer, evaluation incentive adjustment layer and output layer are integrated to obtain a complete energy computing power interconnection model. Through the construction of the energy computing power interconnection model, the power dispatching simulation condition data set can be input into the model for evaluation and analysis, so as to more efficiently obtain a prediction quality coefficient that can reflect the quality of the simulated dispatching.

[0053] Furthermore, the method further comprises:

[0054] If the first decision prediction quality coefficient does not meet the predetermined decision quality constraint, eliminating the first power dispatch decision;

[0055] Continue to randomly extract the power dispatch decision space to obtain a second power dispatch decision;

[0056] Calculating a second decision prediction quality coefficient corresponding to the second power dispatch decision according to the energy computing power interconnection model;

[0057] Determining whether the second decision prediction quality coefficient satisfies the predetermined decision quality constraint;

[0058] If the second decision prediction quality coefficient satisfies the predetermined decision quality constraint, adding the second power dispatch decision to the power dispatch decision optimization domain;

[0059] Based on the predetermined decision quality constraint and the energy computing power interconnection model, the power dispatch decision space is continuously iterated and optimized until the power dispatch decision optimization domain that satisfies the predetermined optimization convergence constraint is generated.

[0060] First, according to the predetermined decision quality constraint, the energy computing power interconnection model is used to evaluate the first power dispatch decision, and the corresponding first decision prediction quality coefficient is calculated. If the first decision prediction quality coefficient does not meet the predetermined decision quality constraint, the decision is eliminated and the next step is entered. The power dispatch decision space is randomly extracted to obtain the second power dispatch decision, and the second power dispatch decision is evaluated by the energy computing power interconnection model, and the corresponding second decision prediction quality coefficient is calculated, and it is judged whether the second decision prediction quality coefficient meets the predetermined decision quality constraint. If it meets, the second power dispatch decision is added to the power dispatch decision optimization domain; if it does not meet, the power dispatch decision space is randomly extracted again. Based on the predetermined decision quality constraint and the energy computing power interconnection model, the power dispatch decision space is continued to be iteratively optimized. In each iteration, the power dispatch decision is randomly extracted and evaluated, and the decision that meets the predetermined decision quality constraint is added to the power dispatch decision optimization domain. When the power dispatch decision optimization domain meets the predetermined optimization convergence constraint, the iteration is stopped and the optimization result is output. Based on the predetermined decision quality constraints and the energy computing power interconnection model, iterative optimization of the power dispatch decision space can improve the efficiency of dispatch decisions and ultimately obtain a better power dispatch plan.

[0061] Embodiment 2

[0062] Based on the same inventive concept as the power dispatching method based on energy computing power interconnection in the aforementioned embodiment, Figure 2 As shown, the present application provides a power dispatching system based on energy computing power interconnection, the system comprising:

[0063] Prediction feature time zone module 11: The prediction feature time zone module 11 is used to read the prediction feature time zone and the prediction feature environment data corresponding to the prediction feature time zone;

[0064] Load forecasting result module 12: The load forecasting result module 12 is used to perform load forecasting on the target power system according to the forecasting characteristic time zone and the forecasting characteristic environment data to obtain a target load forecasting result;

[0065] Dispatching decision space module 13: the dispatching decision space module 13 is used to perform power dispatching decision mining on the target power system based on the target load forecast result to obtain a power dispatching decision space;

[0066] Optimization and convergence constraint module 14: The optimization and convergence constraint module 14 is used to iteratively optimize the power dispatch decision space based on a predetermined decision quality constraint and a predetermined optimization convergence constraint according to a pre-built energy computing power interconnection model to obtain a power dispatch decision optimization domain;

[0067] Deep optimization analysis module 15: The deep optimization analysis module 15 is used to perform deep optimization analysis based on the power dispatch decision optimization domain to obtain a power dispatch solution;

[0068] Power system dispatching module 16: The power system dispatching module 16 is used to perform power dispatching on the target power system based on the power dispatching plan.

[0069] Furthermore, the load forecasting result module 12 further includes:

[0070] Obtaining a load record data set of the target power system;

[0071] Performing principal component analysis on the load record data set to obtain a dimension-reduced load record data set;

[0072] Based on the Markov chain, data integration is performed on the dimension-reduced load record data set to obtain a load forecast chain;

[0073] The prediction characteristic time zone and the prediction characteristic environment data are input into the load prediction chain to obtain the target load prediction result.

[0074] Furthermore, the scheduling decision space module 13 also includes:

[0075] Reading the power dispatch decision record of the target power system according to the target load forecast result to obtain a power dispatch decision record set;

[0076] Calibrate the trigger interval of the power dispatch decision indicator based on the power dispatch decision record set to obtain multiple power dispatch decision indicator trigger intervals;

[0077] Traversing the multiple power dispatch decision indicator trigger intervals to perform multiple random value acquisitions to obtain multiple power dispatch decisions;

[0078] The plurality of power dispatch decisions are added to the power dispatch decision space.

[0079] Furthermore, the optimization convergence constraint module 14 further includes:

[0080] Performing random extraction according to the power dispatch decision space to obtain a first power dispatch decision;

[0081] Calculating a first decision prediction quality coefficient corresponding to the first power dispatch decision according to the energy computing power interconnection model;

[0082] Determining whether the first decision prediction quality coefficient satisfies the predetermined decision quality constraint;

[0083] If the first decision prediction quality coefficient satisfies the predetermined decision quality constraint, adding the first power dispatch decision to the power dispatch decision optimization domain;

[0084] Based on the predetermined decision quality constraint and the energy computing power interconnection model, the power dispatch decision space is continuously iterated and optimized until the power dispatch decision optimization domain that satisfies the predetermined optimization convergence constraint is generated.

[0085] Furthermore, the optimization convergence constraint module 14 further includes:

[0086] Modeling the target power system based on the digital twin platform to obtain a power simulation model;

[0087] According to the first power dispatching decision, the power simulation model is simulated for power dispatching to obtain a first decision-simulated operating condition data set;

[0088] The energy computing power interconnection model includes an input layer, a simulation scheduling evaluation layer, an evaluation incentive adjustment layer and an output layer;

[0089] Inputting the first decision simulation working condition data set into the simulation scheduling evaluation layer to obtain a first simulation scheduling evaluation result, wherein the simulation scheduling evaluation layer includes multi-dimensional simulation scheduling evaluation indicators, and the multi-dimensional simulation scheduling evaluation indicators include power scheduling safety, power scheduling energy utilization, power scheduling load adaptability and power scheduling stability;

[0090] The first simulation scheduling evaluation result is input into the evaluation incentive adjustment layer to obtain the first decision prediction quality coefficient, wherein the evaluation incentive adjustment layer includes preset evaluation incentive adjustment constraints.

[0091] Furthermore, the optimization convergence constraint module 14 further includes:

[0092] If the first decision prediction quality coefficient does not meet the predetermined decision quality constraint, eliminating the first power dispatch decision;

[0093] Continue to randomly extract the power dispatch decision space to obtain a second power dispatch decision;

[0094] Calculating a second decision prediction quality coefficient corresponding to the second power dispatch decision according to the energy computing power interconnection model;

[0095] Determining whether the second decision prediction quality coefficient satisfies the predetermined decision quality constraint;

[0096] If the second decision prediction quality coefficient satisfies the predetermined decision quality constraint, adding the second power dispatch decision to the power dispatch decision optimization domain;

[0097] Based on the predetermined decision quality constraint and the energy computing power interconnection model, the power dispatch decision space is continuously iterated and optimized until the power dispatch decision optimization domain that satisfies the predetermined optimization convergence constraint is generated.

[0098] Furthermore, the deep optimization analysis module 15 also includes:

[0099] According to the power dispatch decision optimization domain, a first winning power dispatch decision and a first winning decision prediction quality coefficient corresponding to the first winning power dispatch decision are selected;

[0100] Randomly adjusting each power dispatching decision in the power dispatching decision optimization domain to obtain multiple power dispatching decision neighborhoods;

[0101] Based on the energy computing power interconnection model, calculating multiple neighborhood-decision prediction quality coefficients corresponding to the multiple power dispatch decision neighborhoods;

[0102] Based on the multiple neighborhood-decision prediction quality coefficients and the multiple power dispatch decision neighborhoods, screening the optimal neighborhood-decision prediction quality coefficient and the optimal neighborhood-power dispatch decision corresponding to the optimal neighborhood-decision prediction quality coefficient;

[0103] Determining whether the optimal neighborhood-decision prediction quality coefficient is greater than the first winning decision prediction quality coefficient;

[0104] If the optimal neighborhood-decision prediction quality coefficient is greater than the first winning decision prediction quality coefficient, outputting the optimal neighborhood-power dispatch decision as the power dispatch plan;

[0105] If the optimal neighborhood-decision prediction quality coefficient is less than / equal to the first winning decision prediction quality coefficient, the first winning power dispatching decision is output as the power dispatching plan.

[0106] Through the above-mentioned detailed description of the power dispatching method based on energy computing power interconnection in this specification, those skilled in the art can clearly understand the power dispatching method based on energy computing power interconnection in this embodiment. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0107] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A power dispatching method based on energy computing power interconnection, characterized in that: The method comprises: Reading the predicted characteristic time zone and the predicted characteristic environment data corresponding to the predicted characteristic time zone; Perform load forecasting on the target power system according to the forecast characteristic time zone and the forecast characteristic environmental data to obtain a target load forecast result; Based on the target load forecast result, power dispatch decision mining is performed on the target power system to obtain a power dispatch decision space; Based on the predetermined decision quality constraint and the predetermined optimization convergence constraint, the power dispatch decision space is iteratively optimized according to the pre-built energy computing power interconnection model to obtain the power dispatch decision optimization domain; Performing in-depth optimization analysis based on the power dispatch decision optimization domain to obtain a power dispatch plan; The target power system is subjected to power dispatch based on the power dispatch plan.

2. The method according to claim 1, characterized in that Performing load forecasting on a target power system according to the forecast characteristic time zone and the forecast characteristic environment data to obtain a target load forecasting result includes: Obtaining a load record data set of the target power system; Performing principal component analysis on the load record data set to obtain a dimension-reduced load record data set; Based on the Markov chain, data integration is performed on the dimension-reduced load record data set to obtain a load forecast chain; The prediction characteristic time zone and the prediction characteristic environment data are input into the load prediction chain to obtain the target load prediction result.

3. The method according to claim 1, characterized in that Based on the target load forecast result, power dispatch decision mining is performed on the target power system to obtain a power dispatch decision space, including: Reading the power dispatch decision record of the target power system according to the target load forecast result to obtain a power dispatch decision record set; Calibrate the trigger interval of the power dispatch decision indicator based on the power dispatch decision record set to obtain multiple power dispatch decision indicator trigger intervals; Traversing the multiple power dispatch decision indicator triggering intervals to perform multiple random value acquisitions to obtain multiple power dispatch decisions; The plurality of power dispatch decisions are added to the power dispatch decision space.

4. The method according to claim 1, characterized in that Based on the predetermined decision quality constraint and the predetermined optimization convergence constraint, the power dispatch decision space is iteratively optimized according to the pre-built energy computing power interconnection model to obtain the power dispatch decision optimization domain, including: Performing random extraction according to the power dispatch decision space to obtain a first power dispatch decision; Calculating a first decision prediction quality coefficient corresponding to the first power dispatch decision according to the energy computing power interconnection model; Determining whether the first decision prediction quality coefficient satisfies the predetermined decision quality constraint; If the first decision prediction quality coefficient satisfies the predetermined decision quality constraint, adding the first power dispatch decision to the power dispatch decision optimization domain; Based on the predetermined decision quality constraint and the energy computing power interconnection model, the power dispatch decision space is continuously iterated and optimized until the power dispatch decision optimization domain that satisfies the predetermined optimization convergence constraint is generated.

5. The method according to claim 4, characterized in that Calculating a first decision prediction quality coefficient corresponding to the first power dispatch decision according to the energy computing power interconnection model includes: Modeling the target power system based on the digital twin platform to obtain a power simulation model; According to the first power dispatching decision, the power simulation model is simulated for power dispatching to obtain a first decision-simulated operating condition data set; The energy computing power interconnection model includes an input layer, a simulation scheduling evaluation layer, an evaluation incentive adjustment layer and an output layer; Inputting the first decision simulation working condition data set into the simulation scheduling evaluation layer to obtain a first simulation scheduling evaluation result, wherein the simulation scheduling evaluation layer includes multi-dimensional simulation scheduling evaluation indicators, and the multi-dimensional simulation scheduling evaluation indicators include power scheduling safety, power scheduling energy utilization, power scheduling load adaptability and power scheduling stability; The first simulation scheduling evaluation result is input into the evaluation incentive adjustment layer to obtain the first decision prediction quality coefficient, wherein the evaluation incentive adjustment layer includes preset evaluation incentive adjustment constraints.

6. The method according to claim 4, characterized in that Determining whether the first decision prediction quality coefficient satisfies the predetermined decision quality constraint further includes: If the first decision prediction quality coefficient does not meet the predetermined decision quality constraint, eliminating the first power dispatch decision; Continue to randomly extract the power dispatch decision space to obtain a second power dispatch decision; Calculating a second decision prediction quality coefficient corresponding to the second power dispatch decision according to the energy computing power interconnection model; Determining whether the second decision prediction quality coefficient satisfies the predetermined decision quality constraint; If the second decision prediction quality coefficient satisfies the predetermined decision quality constraint, adding the second power dispatch decision to the power dispatch decision optimization domain; Based on the predetermined decision quality constraint and the energy computing power interconnection model, the power dispatch decision space is continuously iterated and optimized until the power dispatch decision optimization domain that satisfies the predetermined optimization convergence constraint is generated.

7. The method according to claim 1, characterized in that Based on the power dispatch decision optimization domain, a deep optimization analysis is performed to obtain a power dispatch solution, including: According to the power dispatch decision optimization domain, a first winning power dispatch decision and a first winning decision prediction quality coefficient corresponding to the first winning power dispatch decision are selected; Randomly adjusting each power dispatching decision in the power dispatching decision optimization domain to obtain multiple power dispatching decision neighborhoods; Based on the energy computing power interconnection model, calculating multiple neighborhood-decision prediction quality coefficients corresponding to the multiple power dispatch decision neighborhoods; Based on the multiple neighborhood-decision prediction quality coefficients and the multiple power dispatch decision neighborhoods, screening the optimal neighborhood-decision prediction quality coefficient and the optimal neighborhood-power dispatch decision corresponding to the optimal neighborhood-decision prediction quality coefficient; Determining whether the optimal neighborhood-decision prediction quality coefficient is greater than the first winning decision prediction quality coefficient; If the optimal neighborhood-decision prediction quality coefficient is greater than the first winning decision prediction quality coefficient, outputting the optimal neighborhood-power dispatch decision as the power dispatch plan; If the optimal neighborhood-decision prediction quality coefficient is less than / equal to the first winning decision prediction quality coefficient, the first winning power dispatching decision is output as the power dispatching plan.

8. The power dispatching system based on energy computing power interconnection is characterized by: The system comprises: Prediction feature time zone module: reads the prediction feature time zone and the prediction feature environment data corresponding to the prediction feature time zone; Load forecasting result module: performs load forecasting on the target power system according to the forecasting characteristic time zone and the forecasting characteristic environment data to obtain a target load forecasting result; Dispatching decision space module: based on the target load forecast result, the target power system is subjected to power dispatch decision mining to obtain the power dispatch decision space; Optimization convergence constraint module: based on the predetermined decision quality constraint and the predetermined optimization convergence constraint, the power dispatch decision space is iteratively optimized according to the pre-built energy computing power interconnection model to obtain the power dispatch decision optimization domain; Deep optimization analysis module: performs deep optimization analysis based on the power dispatch decision optimization domain to obtain a power dispatch solution; Power system dispatching module: performs power dispatching on the target power system based on the power dispatching plan.

Citation Information

Patent Citations

  • Power system simulation scheduling method and system based on reinforcement learning

    CN114139354A

  • Electric power energy intelligent scheduling method and system based on digital twinning

    CN117833242A

  • Energy coordination optimization control method and system for virtual power plant

    CN118228887A

Cited By

  • Intelligent control method and system for multi-tower desulfurization spraying device

    CN119327244A