Virtual power plant load resource regulation and control method and system

Abstract

The invention discloses a virtual power plant load resource regulation and control method and system, and the method comprises the steps: carrying out the knowledge extraction of a virtual power plant based on a virtual power plant knowledge base through a knowledge extraction strategy, and obtaining the data of the virtual power plant; determining a virtual power plant operation strategy according to the virtual power plant data; determining a regulation and control economy evaluation index, a regulation and control power supply capability evaluation index and a regulation and control credit index according to the virtual power plant operation strategy; performing resource regulation capability evaluation based on the regulation economy evaluation index, the regulation power supply capability evaluation index and the regulation credit index to obtain regulation data; and performing virtual power plant load resource regulation and control according to the regulation and control data. The problems of low efficiency and poor accuracy of virtual power plant resource regulation and control are solved, the load resource regulation and control efficiency is improved, and the regulation and control accuracy is high.

Description

Technical Field

[0001] The present invention relates to the technical field of resource regulation, and in particular to a method and system for regulating load resources of a virtual power plant. Background Art

[0002] With the introduction of energy sector carbon reduction targets and the development of new power systems, virtual power plants (VPPs) are playing a key role in increasing power system flexibility by aggregating distributed energy resources. However, the large number and dispersed distribution of distributed resources, such as distributed power sources, energy storage, and flexible loads, makes resource allocation difficult and necessitates scheduling. However, current VPP resource control methods are inefficient and inaccurate, making it difficult to meet production needs.

[0003] In the prior art, an invention patent with the publication number CN118410982A discloses a method and system for optimizing the economic dispatch of a virtual power plant. The method comprises: obtaining the economic cost of the virtual power plant, which includes the cost of the virtual power plant participating in the market purchase and sale of electricity, the start-up and shutdown costs and operating costs of the gas turbine, the loss costs of electric vehicles, and the compensation costs of load reduction; constructing an objective function based on the economic cost; obtaining constraints, which include gas turbine constraints, battery constraints, interruptible load constraints, central air conditioning system constraints, electric vehicle constraints, and energy balance constraints; and performing economic dispatch based on the objective function and the constraints. However, the dispatching scheme of this patent directly constructs the objective function and constraints based on economic-related parameters, and then performs economic dispatch, resulting in poor control accuracy. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to solve the technical problem of low efficiency and poor accuracy in resource control of virtual power plants in the prior art.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] A virtual power plant load resource control method, comprising:

[0007] Based on the virtual power plant knowledge base, virtual power plant knowledge is extracted through knowledge extraction strategy to obtain virtual power plant data;

[0008] Determine the virtual power plant operation strategy based on virtual power plant data;

[0009] Determine the control economic evaluation index, control power supply capacity evaluation index and control credit index based on the virtual power plant operation strategy;

[0010] Evaluate resource regulation capabilities based on regulation economic evaluation indicators, regulation power supply capacity evaluation indicators, and regulation credit indicators to obtain regulation data;

[0011] The load resources of the virtual power plant are regulated based on the regulation data.

[0012] In one embodiment of the present invention, the knowledge extraction strategy includes a first knowledge extraction strategy and a second knowledge extraction strategy;

[0013] Extracting related knowledge from the virtual power plant knowledge base through the first knowledge extraction strategy, the related knowledge includes electricity price and load demand;

[0014] Extracting prediction knowledge from the virtual power plant knowledge base through the second knowledge extraction strategy, the prediction knowledge includes predicted power consumption, predicted power generation and predicted load;

[0015] The electricity price, load demand, predicted electricity consumption, predicted power generation and predicted load are used as virtual power plant data.

[0016] In one embodiment of the present invention, extracting prediction knowledge includes:

[0017] Knowledge integration is performed based on the virtual power plant knowledge base to obtain integrated knowledge; a knowledge graph is constructed based on the integrated knowledge; attribute extraction, relationship extraction and entity extraction are performed based on the knowledge graph to obtain attributes, relationships and entities; prediction is performed based on the attributes, relationships and entities through a graph neural network model to obtain prediction knowledge.

[0018] In one embodiment of the present invention, determining the regulation economic evaluation index, the regulation power supply capacity evaluation index, and the regulation credit index includes:

[0019] Determine the control economic evaluation index based on the electricity price and the preset dispatch economic coefficient;

[0020] The virtual power plant operation strategy determines the number of resource dispatches, the amount of resource dispatches, the remaining grid connection time, the total grid connection time, the current dispatchable power, and the dispatchable power threshold;

[0021] Determine the evaluation index of power supply control capability based on the remaining grid connection time, total grid connection time, current dispatchable power and dispatchable power threshold;

[0022] Determine the control credit index based on the number of resource scheduling and the amount of resource scheduling.

[0023] In one embodiment of the present invention, the formula for regulating the economic evaluation index is:

[0024]

[0025] Where, J m (t) represents the control economic evaluation index of resource m in time period t, A represents the preset scheduling economic coefficient, k 3m It is represented as a fixed value related to resource m;

[0026] The formula for evaluating the power supply capability is:

[0027]

[0028] Where G m (t) represents the evaluation index of the power supply capacity of resource m in time period t, T1(t) represents the remaining grid connection time, T2(t) represents the total grid connection time, P m,max (t), P max (t) represents the schedulable power of resource m at the current moment and the maximum schedulable power in the resource, respectively;

[0029] The formula for regulating the credit index is:

[0030]

[0031] Where, X m (t) represents the control credit index of resource m in time period t, K m It represents the number of resource scheduling, k represents the kth resource scheduling, P1 represents the actual scheduling amount of resource m, and P2 represents the expected scheduling amount of resource m.

[0032] In one embodiment of the present invention, obtaining control data includes:

[0033] Obtain the load peak-valley difference data and the consumption capacity data of each load resource of the virtual power plant, and determine the objective function based on the load peak-valley difference data and the new energy consumption capacity data;

[0034] Construct operational safety constraints, new energy operational constraints, and resource operational constraints, and use operational safety constraints, new energy operational constraints, and resource operational constraints as constraint conditions;

[0035] Construct a regulatory capacity evaluation model based on the objective function and constraints;

[0036] Based on the regulation economic evaluation index, regulation power supply capacity evaluation index and regulation credit index, the regulation capacity of each resource in the virtual power plant is evaluated respectively through the regulation capacity evaluation model to obtain the regulation data.

[0037] In one embodiment of the present invention, the objective function is the following formula:

[0038]

[0039] Where, L i,t is the net load of the node, [T1, T2] is the time period to be evaluated, where T1 is the starting time, T2 is the ending time, n is the number of distribution network nodes, C j,tis the consumption of the j-th renewable energy source at time t, and m is the number of renewable energy sources connected.

[0040] In one embodiment of the present invention, performing load resource control of a virtual power plant according to control data includes:

[0041] Determine the target virtual power plant load resources based on the control data;

[0042] Obtain energy data of the target virtual power plant load resources, and aggregate the virtual power plant resources based on the energy data to obtain aggregated data;

[0043] Conduct virtual power plant value analysis based on aggregated data and obtain analysis results;

[0044] The load resources of virtual power plants are regulated based on the analysis results.

[0045] In one embodiment of the present invention, the virtual power plant knowledge base is obtained by:

[0046] Collecting virtual power plant flow data and virtual power plant historical data, and preprocessing the virtual power plant flow data and virtual power plant historical data to obtain preprocessed data;

[0047] Build a knowledge extraction rule base;

[0048] Generate virtual power plant knowledge units based on the rules in the knowledge extraction rule base and pre-processed data;

[0049] Generate a virtual power plant knowledge base based on virtual power plant knowledge units.

[0050] The present invention further provides a virtual power plant load resource control system, which applies the above-mentioned virtual power plant load resource control method and includes:

[0051] An extraction module is used to extract virtual power plant knowledge based on the virtual power plant knowledge base through a knowledge extraction strategy to obtain virtual power plant data;

[0052] An operation strategy module, used to determine the virtual power plant operation strategy based on the virtual power plant data;

[0053] An indicator module is used to determine the control economy evaluation index, the control power supply capacity evaluation index and the control credit index according to the virtual power plant operation strategy;

[0054] An evaluation module is used to evaluate resource regulation capabilities based on regulation economic evaluation indicators, regulation power supply capacity evaluation indicators, and regulation credit indicators, and obtain regulation data;

[0055] The control module is used to control the load resources of the virtual power plant according to the control data.

[0056] Compared with the existing technology, the beneficial effects of the present invention are: through the comprehensive analysis of the regulation of economic evaluation indicators, the regulation of power supply capacity evaluation indicators and the regulation of credit indicators, the virtual power plant can achieve optimal resource allocation, high regulation accuracy, improve overall operating efficiency, reduce operating costs, and ultimately achieve the purpose of improving economic benefits and ensuring power supply stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flow chart of a virtual power plant load resource control method according to an embodiment of the present invention.

[0058] Figure 2 This is a flowchart of obtaining a virtual power plant knowledge base according to an embodiment of the present invention.

[0059] Figure 3 This is a block diagram of a virtual power plant load resource control system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] To facilitate those skilled in the art to understand the technical solution of the present invention, the technical solution of the present invention is further described with reference to the accompanying drawings.

[0061] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature specified as "first" or "second" may explicitly or implicitly include one or more of such features. Throughout the description of this application, "plurality" means two or more, unless otherwise specifically defined.

[0062] See also Figure 1 As shown, the present invention provides a virtual power plant load resource control method, comprising:

[0063] S10, extracting virtual power plant knowledge based on the virtual power plant knowledge base through a knowledge extraction strategy to obtain virtual power plant data.

[0064] In one embodiment of the present invention, the execution subject of this embodiment is a virtual power plant load resource control device, and can also be other devices with the same or similar functions. This embodiment does not impose specific restrictions on this. This embodiment is illustrated by taking the virtual power plant load resource control device as an example.

[0065] In this embodiment, the virtual power plant knowledge base is a structured collection that stores information related to the virtual power plant. Through the knowledge extraction strategy, data related to the virtual power plant, namely virtual power plant data, can be extracted from the virtual power plant knowledge base.

[0066] In this embodiment, the knowledge extraction strategy includes a first knowledge extraction strategy and a second knowledge extraction strategy. The knowledge extraction strategy is used to extract virtual power plant knowledge from the virtual power plant knowledge base to obtain virtual power plant data. The strategy includes: extracting associated knowledge from the virtual power plant knowledge base using the first knowledge extraction strategy. The associated knowledge includes electricity prices and load demands. The second knowledge extraction strategy is used to extract predictive knowledge from the virtual power plant knowledge base. The predictive knowledge includes predicted electricity consumption, predicted power generation, and predicted load. The electricity price, load demand, predicted electricity consumption, predicted power generation, and predicted load are used as virtual power plant data.

[0067] In this embodiment, in the context of a virtual power plant (VPP), the knowledge extraction strategy is based on the understanding of the power system and demand forecast, and aims to extract relevant information from the virtual power plant knowledge base to support power dispatch, optimization and forecasting tasks. The first knowledge extraction strategy focuses on mining the correlation between the elements in the virtual power plant knowledge base, and extracts key information by analyzing the correlation in the virtual power plant historical data and real-time streaming data. The second knowledge extraction strategy focuses on mining future trends and patterns from the virtual power plant knowledge base. The virtual power plant knowledge base stores a large amount of virtual power plant historical data and real-time streaming data, which is convenient for direct analysis based on the data without the need for operations such as data calls, thereby improving processing efficiency.

[0068] In this embodiment, virtual power plant knowledge refers to both virtual power plant operational data and information and previously unknown, valid, and applicable concepts, rules, patterns, laws, and constraints contained in the massive, incomplete, noisy, fuzzy, and random real-world power data. Virtual power plant knowledge can be categorized into two types: associative knowledge and predictive knowledge, depending on the underlying information.

[0069] In this embodiment, the associated knowledge in the virtual power plant reflects the dependencies or associations between one event and other events. If two or more attributes of two events are associated, the value of one attribute can be predicted based on the values ​​of the others. This associated knowledge primarily includes electricity prices and load demand, as well as knowledge about power system stability analysis, power system fault diagnosis, and power system anomaly monitoring.

[0070] In this embodiment, predictive knowledge is extracted from the virtual power plant knowledge base using a second knowledge extraction strategy, including: performing knowledge integration based on the virtual power plant knowledge base to obtain integrated knowledge; constructing a knowledge graph based on the integrated knowledge; performing attribute extraction, relationship extraction, and entity extraction based on the knowledge graph to obtain attributes, relationships, and entities; and performing predictions based on the attributes, relationships, and entities using a graph neural network model to obtain predictive knowledge.

[0071] In this embodiment, knowledge integration coordinates, combines, and unifies knowledge from different sources to form a more comprehensive, consistent, and operational knowledge system. Knowledge integration is performed based on the relationships between knowledge units in the virtual power plant knowledge base. The integrated knowledge is then graphically represented to showcase the relationships, hierarchy, and semantics of the knowledge, forming a knowledge graph. A knowledge graph is a structured semantic knowledge base whose basic unit is an "entity-relationship-entity" triple.

[0072] In this embodiment, the basic structure of the knowledge graph is constructed based on the integrated data. The basic structure of the knowledge graph includes:

[0073] Entity: The representative of each data point or data set. Entities are usually power plants, power generation equipment, load demand, predicted power consumption, etc.

[0074] Attribute: A characteristic or description of an entity. For example, the attributes of a power plant can be its name, geographical location, and power generation capacity; the attributes of load demand can be hourly electricity consumption, demand type, etc.

[0075] Relationship: A connection or dependency between entities, for example, a power plant "generates" electricity, load demand "affects" electricity prices, and power generation "predicts" load demand.

[0076] Based on this information, a directed graph structure is formed, where entities and relationships are connected by edges.

[0077] The initial information of the entity (such as the power plant's power generation capacity, electricity price, load demand, etc.) is input into the Graph Neural Network (GNN) model as node features. The GNN model aggregates the node's neighbor information through multi-layer graph convolution or graph attention mechanism, updates the representation of each node layer by layer, and predicts future electricity consumption, load, power generation and other information based on the learning of nodes and edges through the output layer of the graph neural network.

[0078] In this embodiment, the GNN model (Graph Neural Network) is a deep learning model for processing graph-structured data. In the virtual power plant knowledge graph, using the GNN model to make predictions based on attributes, relationships, and entities requires converting the knowledge graph data into a format suitable for the GNN model, where nodes and edges are represented as feature matrices and adjacency matrices.

[0079] In another embodiment of the present invention, a method for obtaining a virtual power plant knowledge base is provided, comprising:

[0080] S101 , collecting virtual power plant flow data and virtual power plant historical data, and preprocessing the virtual power plant flow data and virtual power plant historical data to obtain preprocessed data.

[0081] In this embodiment, virtual power plant flow data refers to data generated in real time or near real time, such as current load demand, real-time power generation of distributed energy, electricity market price changes, etc.

[0082] Virtual power plant historical data refers to the virtual power plant operation data collected over a period of time in the past, such as historical load demand, historical electricity prices, historical power generation, etc.

[0083] In this embodiment, real-time acquisition of virtual power plant streaming data to ensure data timeliness can be accomplished through real-time sensor data, API calls, or other methods, which are not specifically limited in this embodiment. Acquisition of virtual power plant historical data can be accomplished through database queries, log file imports, and other methods to ensure the integrity and accuracy of historical data, which are not specifically limited in this embodiment.

[0084] In this embodiment, virtual power plant flow data and virtual power plant historical data are collected and preprocessed to obtain preprocessed data, including:

[0085] Real-time streaming data from the virtual power plant is collected and merged using a sliding window model to generate virtual power plant stream data. Transformations are performed on this virtual power plant stream data to generate stream data features. Stream data features are representative statistics or indicators extracted from the virtual power plant stream data. These features can reflect the real-time operating status and trends of the virtual power plant, such as load demand, power generation efficiency, and equipment failure rate.

[0086] Acquire historical data from the virtual power plant, perform data processing and discretization on the historical data, and obtain historical data features. The data processing includes at least data extraction, data conversion, and data loading. The stream data features and historical data features are used as preprocessed data.

[0087] In this embodiment, the stream data is the real-time data involved in the virtual power plant. Let t represent any timestamp, a t Represents the data arriving at this timestamp. The data stream can be represented as {…a t-1 ,a t ,a t+1 …}.

[0088] It is understandable that the sliding window parameters are determined, including the size of the sliding window and the sliding step size, wherein the window size determines the amount of data processed each time, and the sliding step size determines the degree of overlap between windows. According to the sliding window parameters, real-time data collection is performed in the virtual power plant, and the data in the sliding window is merged to generate an overall data set. The merging can be performed by simple summation, average value, etc., depending on the type and requirements of the data, and this embodiment does not impose specific restrictions on this. Sliding the window forward by a set step size to prepare for processing the next set of data can maintain continuous monitoring of the real-time streaming data, and ultimately obtain the virtual power plant streaming data. The specific sliding window model is not limited in this embodiment.

[0089] It is worth mentioning that the purpose of extracting features from virtual power plant flow data is to better understand the patterns and trends of the data. Wavelet transform is performed on the virtual power plant flow data, that is, the time domain data is converted into the time-frequency domain, the instantaneous frequency of the signal is obtained, and the flow data features are obtained.

[0090] S102, building a knowledge extraction rule base.

[0091] In one embodiment of the present invention, the knowledge extraction rule base includes decision rules and custom rules. Decision rules primarily include algorithms, models, and formulas commonly used in extracting power forecasting knowledge and related knowledge, and can be entered based on actual circumstances. Custom rules are rules that users enter as needed based on the rules defined in the rule base to suit their specific circumstances.

[0092] It can be understood that the decision rules at least include load forecasting rules, fault diagnosis rules, system status analysis rules and market demand analysis rules.

[0093] S103: Generate a virtual power plant knowledge unit based on the rules in the knowledge extraction rule base and the preprocessed data.

[0094] In one embodiment of the present invention, key information is extracted from the preprocessed data based on the rules in the knowledge extraction rule base, the key information is organized into knowledge units, and the knowledge units are formatted into a suitable structure to obtain virtual power plant knowledge units for easy storage, query and analysis. JSON, XML and other formats can be used, and this embodiment does not impose any specific restrictions on this.

[0095] In the specific implementation, before organizing key information into knowledge units, it also includes: verifying the key information to ensure its accuracy and consistency, performing necessary data quality control, and handling possible errors or exceptions.

[0096] It is worth noting that after generating a virtual power plant knowledge unit, you also need to add documentation and annotations to the generated virtual power plant knowledge unit, explaining the meaning, calculation method, and data source of each indicator. This will facilitate subsequent use and understanding.

[0097] In this embodiment, the virtual power plant knowledge unit includes an operating status knowledge unit, an energy production knowledge unit, a load demand knowledge unit, a power dispatching and optimization knowledge unit, and a market transaction knowledge unit.

[0098] Operation status knowledge unit: describes the operation status of each device or system in the virtual power plant, such as the working status of the power generation equipment, the charging and discharging status of the battery storage, the load fluctuation, etc.

[0099] Energy production knowledge unit: related to the energy production process in the virtual power plant, such as power generation, power generation type (solar energy, wind energy, thermal power, etc.), power generation efficiency, predicted power generation capacity, etc.

[0100] Load demand knowledge unit: involves changes in load demand of virtual power plants, such as real-time power demand, load forecasting, demand response, etc.

[0101] Power dispatch and optimization knowledge unit: Knowledge units related to grid dispatch, load optimization, power generation and energy storage dispatch, etc., help optimize power distribution and resource scheduling.

[0102] Fault detection and early warning knowledge unit: Generate fault detection and early warning information based on anomalies or deviations in data streams to promptly reflect possible failure risks of the system or equipment.

[0103] Market transaction knowledge unit: This unit covers electricity prices, market demand, market supply and demand matching in the power market, helping virtual power plants make market transaction decisions.

[0104] Each knowledge unit should also include data source information, feature information, time label, etc.

[0105] S104: Generate a virtual power plant knowledge base based on the virtual power plant knowledge unit.

[0106] In one embodiment of the present invention, the structure of the virtual power plant knowledge base is determined, that is, the organization of knowledge units, such as main categories, subcategories, and associations, etc., which is not specifically limited in this embodiment. The generated virtual power plant knowledge units are integrated into the knowledge base according to the organization of the knowledge units, ensuring that each knowledge unit is correctly classified and establishing appropriate relationship links. The data format in the knowledge base is standardized to ensure data consistency and queryability. An index is established for the data in the knowledge base to enable rapid retrieval and query of relevant information, which helps improve the query efficiency of the knowledge base. Metadata and descriptive information are added to each knowledge unit in the knowledge base, including data source, timestamp, quality control information, etc. Correlation analysis is performed to identify the relationships between different knowledge units and obtain a well-organized database. Specifically, supervised learning (such as decision trees, random forests, and support vector machines) is used to model the relationships between different knowledge units and predict the possible changes in other knowledge units when a certain knowledge unit changes. A suitable database or knowledge storage system is selected and the organized knowledge base is stored therein to form a virtual power plant knowledge base.

[0107] In one embodiment of the present invention, through the above method, a virtual power plant knowledge unit is generated based on the data preprocessed by the rules in the knowledge extraction rule base, thereby generating a virtual power plant knowledge base for knowledge extraction, improving data accuracy, and thus performing more efficient and accurate virtual power plant load resource regulation.

[0108] S20: Determine a virtual power plant operation strategy based on the virtual power plant data.

[0109] In one embodiment of the present invention, an operation strategy is formulated based on key data such as distributed energy resources, electricity market prices, and load demand. The accuracy and real-time performance of virtual power plant data can be improved through knowledge extraction, thereby ensuring the accuracy of the operation strategy.

[0110] In this embodiment, determining a virtual power plant operation strategy based on virtual power plant data includes determining resource scheduling parameters based on load demand, predicted power consumption, predicted power generation, and predicted load. An initial operation strategy is determined based on the resource scheduling parameters and the electricity price. This initial operation strategy is optimized using an operation strategy optimization model to obtain a virtual power plant operation strategy.

[0111] In this embodiment, it should be noted that based on the virtual power plant operation strategy, the virtual power plant system can operate efficiently under different electricity prices and demand conditions.

[0112] In this embodiment, the resource scheduling parameters include the start and stop time of the generator set, the power output, etc. The components of the initial operation strategy include: load and power generation balance and electricity price considerations. Among them, load and power generation balance: preliminarily determine that in a certain time period in the future, the virtual power plant needs to perform resource scheduling based on the difference between predicted load and power generation. Electricity price considerations: in the initial operation strategy, the operation strategy is adjusted according to the electricity price level. For example, when the electricity price is high, efficient and low-cost power generation sources are prioritized. When the electricity price is low, a more flexible and high-cost power generation method may be selected to ensure power supply.

[0113] In this embodiment, in the specific implementation, the electricity price and system operation status are continuously tracked, and the operation strategy is adjusted through the operation strategy optimization model with the support of real-time data. For example:

[0114] The operation strategy is adjusted through the operation strategy optimization model. For example, when the electricity price is low, the charging amount of the energy storage device is increased and the output power of the power generation device is reduced. Conversely, when the electricity price is high, the charging amount of the energy storage device is reduced and the output power of the power generation device is increased. In addition, the operation strategy of the virtual power plant can be adjusted according to the predicted load demand to ensure that there is sufficient power generation capacity to meet the demand during peak load periods, while the power generation can be reduced during low load periods to save costs. In this way, the virtual power plant can flexibly respond to market electricity price fluctuations and load changes to achieve economical and efficient operation. During the optimization process, the operation strategy optimization model will adjust the initial operation strategy based on real-time data, market conditions and demand changes:

[0115] Power generation resource adjustment: Dynamically adjust different power generation methods (such as wind, solar, and traditional thermal power) based on electricity prices, load demand, and power generation forecasts. For example, when electricity prices are high, lower-cost renewable energy can be prioritized, reducing the load on coal-fired or natural gas generators.

[0116] Optimal scheduling of energy storage equipment: Utilize energy storage systems to charge when electricity prices are low and discharge when electricity prices are high, optimizing electricity utilization efficiency and providing flexibility for load fluctuations.

[0117] Load response strategy optimization: Optimize demand response based on load forecast information. When electricity prices are high, demand response can be used to reduce peak load demand and achieve load balancing.

[0118] Optimal scheduling time point selection: adjust the start and stop time of the generator set, optimize the start and stop timing of the unit, reduce startup and maintenance costs, and meet the power supply needs of the power grid.

[0119] S30: Determine a control economy evaluation index, a control power supply capacity evaluation index, and a control credit index according to the virtual power plant operation strategy.

[0120] In one embodiment of the present invention, it should be noted that a virtual power plant is a system that integrates various distributed energy resources (such as wind power, solar power, and energy storage) to achieve the production, scheduling, and trading of electricity through intelligent control and coordinated operations. Determining economic evaluation indicators, power supply capacity evaluation indicators, and credit evaluation indicators for virtual power plants is key to ensuring their sustainable operation.

[0121] In this embodiment, the economic evaluation index is used to measure the difference in resource scheduling costs; the credit evaluation index is used to characterize the completion of resource scheduling within a certain period of time; and the power supply capacity evaluation index is used to quantify the dispatchable potential of large-capacity resources.

[0122] In this embodiment, the control economy evaluation index, the control power supply capacity evaluation index, and the control credit index are determined according to the virtual power plant operation strategy, including:

[0123] Based on electricity prices and preset dispatch economic coefficients, control economic evaluation indicators are determined. The virtual power plant operation strategy determines the number of resource dispatches, resource dispatch volume, remaining grid connection time, total grid connection time, current dispatchable power, and dispatchable power threshold. Based on the remaining grid connection time, total grid connection time, current dispatchable power, and dispatchable power threshold, control power supply capacity evaluation indicators are determined. Based on the number of resource dispatches and resource dispatch volume, control credit indicators are determined.

[0124] In this embodiment, it should be noted that the preset dispatch economic coefficient is a constant that makes the resource economy with the lowest dispatch cost equal to 1, and the economic evaluation index is regulated as shown in the following formula 1:

[0125]

[0126] Where, J m (t) represents the control economic evaluation index of resource m in time period t, A represents the preset scheduling economic coefficient, k 3m It is expressed as a constant value related to resource m. The preset dispatch economic coefficient is a constant set by ensuring that the economy of the lowest dispatch cost resource is 1. The preset dispatch economic coefficient reflects how the system selects the most economically efficient resource in a specific time period, ensuring that resource scheduling can both meet demand and maximize economic benefits.

[0127] The power supply capability evaluation index is as follows:

[0128]

[0129] Where G m(t) represents the evaluation index of the power supply capacity of resource m in time period t, T1(t) represents the remaining grid connection time, T2(t) represents the total grid connection time, P m,max (t), P max (t) represents the dispatchable power of resource m at the current moment and the maximum dispatchable power among the resources, respectively. The concept of grid-connected time refers to the effective period during which a power resource (such as a generator or energy storage device) is connected to the grid and able to provide power. Although these devices typically operate in conjunction with the grid, the grid-connected time division is used to more accurately assess the resource's ability to effectively provide power within a specific time period.

[0130] The resource scheduling amount includes the actual scheduling amount and the expected scheduling amount, and the credit index is regulated as shown in the following formula 3:

[0131]

[0132] Where, X m (t) represents the control credit index of resource m in time period t, K m represents the number of resource dispatches, k represents the kth resource dispatch, P1 represents the actual dispatched capacity of resource m, and P2 represents the expected dispatched capacity of resource m. The regulation credit index is an evaluation of the dispatching entity (such as the grid operator or market operator) on the dispatching execution capability of the virtual power plant. The dispatching entity is concerned with whether the virtual power plant responds to dispatch instructions as expected, whether it can provide sufficient dispatch capacity, and whether it can fulfill its dispatch tasks within the specified time.

[0133] S40, performing resource regulation capability evaluation based on the regulation economic evaluation index, the regulation power supply capability evaluation index, and the regulation credit index, and obtaining regulation data.

[0134] In one embodiment of the present invention, it should be noted that resource regulation capability refers to the ability of a virtual power plant to quickly and flexibly adjust energy production and supply when faced with market changes, load fluctuations or other uncertain factors. It is necessary to comprehensively consider multiple indicators such as scheduling economy, power supply capacity and credit assessment indicators.

[0135] In this embodiment, it can be understood that the dispatching economic evaluation index, the dispatching power supply capacity evaluation index and the dispatching credit evaluation index can quantify the resource regulation capability of the virtual power plant, thereby formulating resource regulation strategies, improving market competitiveness and ensuring contribution to the safety and stability of the power system.

[0136] In this embodiment, resource regulation capability is evaluated based on the regulation economic evaluation index, the regulation power supply capability evaluation index, and the regulation credit index to obtain regulation data, including:

[0137] Obtain peak-to-valley load difference data and consumption data for each load resource in the virtual power plant, and determine an objective function based on this data. Construct operational safety constraints, renewable energy operation constraints, and resource operation constraints, using these constraints as constraints. Based on the objective function and constraints, construct a regulation capability assessment model. Use the regulation capability assessment model to evaluate the regulation capability of each resource in the virtual power plant based on the regulation economic evaluation index, regulation power supply capacity evaluation index, and regulation credit index, and obtain regulation data.

[0138] In this embodiment, the objective function and the constraint conditions constitute a regulation capability evaluation model, and the regulation economic evaluation index, the regulation power supply capability evaluation index, and the regulation credit index serve as inputs to the regulation capability evaluation model.

[0139] In this embodiment, it should be noted that the objective functions are to minimize the load peak-to-valley difference and maximize the new energy consumption, as shown in the following equations 4 and 5:

[0140]

[0141] Where, L i,t is the net load of the node, [T1, T2] is the time period to be evaluated, where T1 is the starting time, T2 is the ending time, n is the number of distribution network nodes, C j,t is the consumption of the j-th renewable energy source at time t, and m is the number of renewable energy sources connected.

[0142] In this embodiment, the objective function in the regulation capability assessment model is chosen to optimize the relevant regulation capability indicators. Therefore, the relationship between the objective function and the regulation economy assessment indicator, the regulation power supply capability assessment indicator, and the regulation credit indicator is reflected through the objective function optimization process. Although the three indicators (regulation economy assessment indicator, regulation power supply capability assessment indicator, and regulation credit indicator) may not be explicitly listed in the objective function expression, they are indirectly reflected through the optimization process.

[0143] Minimizing the load peak-valley difference is directly related to the evaluation indicators of power supply capacity and regulation economy: by minimizing the load peak-valley difference, the volatility of the system can be reduced, thereby improving the stability and economy of power supply.

[0144] Maximizing the consumption of new energy is related to the regulation economic evaluation index and the regulation credit index: maximizing the consumption of new energy helps to improve the utilization rate of clean energy by virtual power plants, reduce the consumption of traditional energy, thereby reducing economic costs, and enhance the power plant's ability to respond to fluctuations in new energy, thereby improving its credit.

[0145] These two objective functions already consider the economic and power supply capabilities associated with regulation capacity. Therefore, the objective functions themselves reflect the influence of these indicators, and the choice of optimization target itself determines how the model balances factors such as regulation economy and power supply capacity.

[0146] The purpose of setting constraints is to ensure that the optimization problem complies with the actual operating limitations during the solution process. These constraints also indirectly reflect the relevant indicators of the regulation capability. Specifically:

[0147] Operational safety constraints: By setting safety limits, we ensure that the grid does not experience overloads or other problems during dispatch. This is closely related to the ability to regulate power supply. A safe and reliable grid naturally has a strong power supply capacity.

[0148] Renewable energy operation constraints: These constraints ensure that the absorption capacity of renewable energy does not exceed its physical limitations and that the absorption capacity of renewable energy is fully utilized in the system's optimal dispatch. This is closely related to the regulation credit and regulation economics. Full utilization of renewable energy can reduce dependence on traditional energy sources, thereby improving economics.

[0149] Resource operational constraints: These constraints help assess the operational capabilities and response times of various resources (such as thermal power and hydropower), ensuring that the regulation process does not exceed the actual capabilities of the resources. This is also evident in the relationship between regulating power supply capacity and regulating economics.

[0150] These constraints not only provide a feasible solution space for the optimization of the objective function, but also affect the final evaluation of the regulation capability. Therefore, they are directly linked to the evaluation indicators related to the regulation capability.

[0151] Although the regulation economy evaluation index, regulation power supply capacity evaluation index, and regulation credit index do not appear directly in explicit mathematical form in the objective function and constraint formulas, the regulation capacity evaluation model itself is an integrated evaluation framework. The three evaluation indicators, as inputs to the regulation capacity evaluation model, can actually be considered as further analysis of the objective function optimization results:

[0152] Economic regulation: When the optimization objectives are to minimize the peak-to-valley difference in load and maximize the consumption of new energy, the system will naturally tend to reduce unnecessary regulation costs, thereby improving economic efficiency.

[0153] Regulating power supply capacity: By ensuring that the difference between load peaks and valleys is minimized, the system can maintain a stable power supply capacity, while maximizing the consumption of new energy contributes to a stable and reliable power supply system.

[0154] Regulation credit: By maximizing the consumption of new energy, the system demonstrates a stronger ability to cope with fluctuations in new energy, thereby enhancing the credit of the virtual power plant.

[0155] These evaluation indicators are evaluated as inputs and are the natural results of the objective function and constraints in the optimization process. They do not need to be listed separately, but are reflected in the solution process of the entire model.

[0156] In the regulation capability assessment model, the objective function and constraints are not static but can be dynamically adjusted based on feedback from evaluation indicators. If significant fluctuations occur in any of the indicators, such as regulation economy, regulation power supply capacity, or regulation credit, during the model's operation, the model can adjust the objective function or constraints to further optimize these aspects.

[0157] In summary, although the evaluation metrics are not explicitly listed in the objective function and constraint formulas, they are indirectly reflected in the optimization process during the objective function and constraint setting. The selection of the objective function and the design of the constraints inherently consider the impact of the three metrics: regulation economy, regulation power supply capacity, and regulation credit. Therefore, these three evaluation metrics are suitable as inputs to the regulation capacity assessment model, and the objective function and constraints in the regulation capacity assessment model effectively account for the impact of these metrics.

[0158] In this embodiment, there are multiple ways to constrain operation safety, new energy operation, and resource operation. This embodiment provides one method for reference:

[0159] In this embodiment, the operational safety constraints are:

[0160]

[0161] Where U i,t 、U min 、u max They are respectively represented as the voltage, voltage lower limit and voltage upper limit at node i in time period t. ij,t 、i ij,max They are respectively represented as the current carrying capacity and safety current of line (i, j) in time period t.

[0162] In this embodiment, the new energy operation constraints are:

[0163]

[0164] Where, They represent the active power output of wind power and photovoltaic power at node i during period t, They are respectively represented as the maximum output power of wind power and photovoltaic power at node i during period t, They are respectively represented as the reactive power output of wind power and photovoltaic power at node i during period t, They are represented as the power factor angles of wind power and photovoltaic power respectively.

[0165] In this embodiment, the resource operation constraints are:

[0166]

[0167] Where, P F,t It is expressed as the actual load of flexible resources on the user side at time t, P F,base,t It is expressed as the benchmark load of flexible resources on the user side at time t, ΔP F,t It is expressed as the adjustment power of the flexible resource on the user side at time t, ΔP F,max,t , ΔP F,min,t They represent the upper and lower adjustment power limits of the user-side flexible resources at time t respectively.

[0168] In this embodiment, by optimizing the objective function and constraints, it can be ensured that the model meets both economic requirements and the reliability of power supply and the effective use of new energy in actual applications. The construction and solution process of the regulation capacity evaluation model usually involves complex mathematical programming and optimization algorithms, such as linear programming, nonlinear programming, integer programming or mixed integer linear programming. These algorithms are capable of handling multi-objective optimization problems and finding the optimal solution while meeting various constraints. Ultimately, the regulation data obtained by the evaluation model can provide a scientific basis for the operation strategy of the virtual power plant, guiding it on how to schedule and manage resources in actual operations, so as to achieve the purpose of improving efficiency, reducing costs and enhancing market competitiveness.

[0169] In this embodiment, linear programming (LP): applicable when all objective functions and constraints are linear, and can be solved by a solver (such as Gurobi, CPLEX).

[0170] Nonlinear programming (NLP): If the objective function or constraints contain nonlinear elements (such as energy storage charging and discharging efficiency, etc.), nonlinear programming methods are required to solve them.

[0171] Integer programming (IP): When the regulation capacity assessment model contains discrete variables (such as generator start and stop status, energy storage charge and discharge control, etc.), integer programming methods can be used to solve it.

[0172] Mixed Integer Linear Programming (MILP): When the regulatory capacity assessment model contains both continuous variables and integer variables, the mixed integer linear programming method can be used to solve it.

[0173] As well, heuristic algorithms and meta-heuristic algorithms: such as genetic algorithms and particle swarm optimization (PSO), are suitable for situations where the solution space is large and the problem is complex.

[0174] S50: Regulate the load resources of the virtual power plant according to the regulation data.

[0175] In one embodiment of the present invention, load resource regulation of a virtual power plant refers to reasonably adjusting various distributed energy resources, energy storage systems, etc. according to market demand, virtual power plant conditions, and system scheduling requirements to meet electricity demand, improve energy utilization efficiency, and ensure system stability.

[0176] In this embodiment, the load resource control of the virtual power plant is performed according to the control data, including:

[0177] Based on the adjustment data, the target virtual power plant load resources are determined, the energy data of the target virtual power plant load resources are obtained, and the virtual power plant resources are aggregated based on the energy data to obtain aggregated data. The virtual power plant value analysis is performed based on the aggregated data to obtain analysis results, and the virtual power plant load resources are regulated based on the analysis results.

[0178] In this embodiment, the virtual power plant value analysis includes distributed energy value analysis, network security verification analysis, and virtual power plant operating cost analysis. The distributed energy value analysis primarily analyzes the interaction between distributed energy sources, grids, loads, and storage to determine the optimal configuration of distributed energy. Network security verification analysis involves calculating network security verification costs under the optimal distributed energy configuration. Virtual power plant operating cost analysis involves calculating the operating costs of the virtual power plant based on the network security verification costs, and regulating the load resources of the virtual power plant with the goal of minimizing the operating costs.

[0179] In this embodiment, virtual power plant value analysis does not conflict with the aforementioned evaluation metrics and objective functions; rather, it serves as a further evaluation and quantification of the optimized solution. Building upon the aforementioned objective functions and evaluation metrics, value analysis primarily transforms virtual power plant operational data into quantitative indicators in economic, environmental, and reliability terms, thereby providing a basis for subsequent scheduling and market decisions. The two are hierarchically related, with the former focusing on model optimization and the latter on the specific value assessment of the optimization results.

[0180] In one embodiment of the present invention, through the above-mentioned method, virtual power plant data is obtained through knowledge extraction and the operation strategy is determined, so as to determine the estimation index to evaluate the resource regulation capability to carry out load resource regulation of the virtual power plant, thereby solving the problems of low efficiency and poor accuracy of virtual power plant resource regulation, improving the load resource regulation efficiency and high regulation accuracy.

[0181] See also Figure 2 As shown, the present invention also provides a virtual power plant load resource control system, which applies the above-mentioned virtual power plant load resource control method, including:

[0182] The extraction module 10 is used to extract virtual power plant knowledge based on the virtual power plant knowledge base through a knowledge extraction strategy to obtain virtual power plant data.

[0183] The operation strategy module 20 is used to determine the virtual power plant operation strategy based on the virtual power plant data.

[0184] The indicator module 30 is used to determine the control economy evaluation index, the control power supply capacity evaluation index and the control credit index according to the virtual power plant operation strategy.

[0185] The evaluation module 40 is used to evaluate the resource regulation capability based on the regulation economic evaluation index, the regulation power supply capability evaluation index and the regulation credit index, and obtain regulation data.

[0186] The control module 50 is used to control the load resources of the virtual power plant according to the control data.

[0187] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description. It is intended that all variations within the meaning and range of equivalents of the claims be embraced herein, and any reference signs in the claims should not be construed as limiting the claims to which they relate.

[0188] The above-mentioned embodiments merely represent the implementation methods of the invention. The protection scope of the present invention is not limited to the above-mentioned embodiments. For those skilled in the art, several variations and improvements can be made without departing from the concept of the present invention, which all fall within the protection scope of the present invention.

Claims

1. A virtual power plant load resource control method, characterized in that: include: Based on the virtual power plant knowledge base, virtual power plant knowledge is extracted through knowledge extraction strategy to obtain virtual power plant data; Determine the virtual power plant operation strategy based on virtual power plant data; Determine the control economic evaluation index, control power supply capacity evaluation index and control credit index based on the virtual power plant operation strategy; Evaluate resource regulation capabilities based on regulation economic evaluation indicators, regulation power supply capacity evaluation indicators, and regulation credit indicators to obtain regulation data; The load resources of the virtual power plant are regulated based on the regulation data.

2. The virtual power plant load resource control method according to claim 1, characterized in that: The knowledge extraction strategy includes a first knowledge extraction strategy and a second knowledge extraction strategy; Extracting related knowledge from the virtual power plant knowledge base through the first knowledge extraction strategy, the related knowledge includes electricity price and load demand; Extracting prediction knowledge from the virtual power plant knowledge base through the second knowledge extraction strategy, the prediction knowledge includes predicted power consumption, predicted power generation and predicted load; The electricity price, load demand, predicted electricity consumption, predicted power generation and predicted load are used as virtual power plant data.

3. The virtual power plant load resource control method according to claim 2, characterized in that: Extract predictive knowledge, including: Knowledge integration is performed based on the virtual power plant knowledge base to obtain integrated knowledge; a knowledge graph is constructed based on the integrated knowledge; attribute extraction, relationship extraction and entity extraction are performed based on the knowledge graph to obtain attributes, relationships and entities; prediction is performed based on the attributes, relationships and entities through a graph neural network model to obtain prediction knowledge.

4. The virtual power plant load resource control method according to claim 1, characterized in that: Determine the control economic evaluation indicators, control power supply capacity evaluation indicators and control credit indicators, including: Determine the control economic evaluation index based on the electricity price and the preset dispatch economic coefficient; The virtual power plant operation strategy determines the number of resource dispatches, the amount of resource dispatches, the remaining grid connection time, the total grid connection time, the current dispatchable power, and the dispatchable power threshold; Determine the evaluation index of power supply control capability based on the remaining grid connection time, total grid connection time, current dispatchable power and dispatchable power threshold; Determine the control credit index based on the number of resource scheduling and the amount of resource scheduling.

5. The virtual power plant load resource control method according to claim 4, characterized in that: The formula for regulating economic evaluation indicators is: Where, J m (t) represents the control economic evaluation index of resource m in time period t, A represents the preset scheduling economic coefficient, k 3m It is represented as a fixed value related to resource m; The formula for evaluating the power supply capability is: Where G m (t) represents the evaluation index of the power supply capacity of resource m in time period t, T1(t) represents the remaining grid connection time, T2(t) represents the total grid connection time, P m,max (t), P max (t) represents the schedulable power of resource m at the current moment and the maximum schedulable power in the resource, respectively; The formula for regulating the credit index is: Where, X m (t) represents the control credit index of resource m in time period t, K m It represents the number of resource scheduling, k represents the kth resource scheduling, P1 represents the actual scheduling amount of resource m, and P2 represents the expected scheduling amount of resource m.

6. The virtual power plant load resource control method according to claim 1, characterized in that: Obtain control data, including: Obtain the load peak-valley difference data and the consumption capacity data of each load resource of the virtual power plant, and determine the objective function based on the load peak-valley difference data and the new energy consumption capacity data; Construct operational safety constraints, new energy operational constraints, and resource operational constraints, and use operational safety constraints, new energy operational constraints, and resource operational constraints as constraint conditions; Construct a regulatory capacity evaluation model based on the objective function and constraints; Based on the regulation economic evaluation index, regulation power supply capacity evaluation index and regulation credit index, the regulation capacity of each resource in the virtual power plant is evaluated respectively through the regulation capacity evaluation model to obtain the regulation data.

7. The virtual power plant load resource control method according to claim 6, characterized in that: The objective function is the following formula: Where, L i,t is the net load of the node, [T1, T2] is the time period to be evaluated, where T1 is the starting time, T2 is the ending time, n is the number of distribution network nodes, C j,t is the consumption of the j-th renewable energy source at time t, and m is the number of renewable energy sources connected.

8. The virtual power plant load resource control method according to claim 1, characterized in that: The load resource regulation of the virtual power plant is carried out based on the regulation data, including: Determine the target virtual power plant load resources based on the control data; Obtain energy data of the target virtual power plant load resources, and aggregate the virtual power plant resources based on the energy data to obtain aggregated data; Conduct virtual power plant value analysis based on aggregated data and obtain analysis results; The load resources of virtual power plants are regulated based on the analysis results.

9. The virtual power plant load resource control method according to claim 1, characterized in that: Access the Virtual Power Plant Knowledge Base through the following methods: Collecting virtual power plant flow data and virtual power plant historical data, and preprocessing the virtual power plant flow data and virtual power plant historical data to obtain preprocessed data; Build a knowledge extraction rule base; Generate virtual power plant knowledge units based on the rules in the knowledge extraction rule base and pre-processed data; Generate a virtual power plant knowledge base based on virtual power plant knowledge units.

10. A virtual power plant load resource control system, characterized in that: The method for controlling load resources of a virtual power plant according to any one of claims 1 to 9 is applied, comprising: An extraction module is used to extract virtual power plant knowledge based on the virtual power plant knowledge base through a knowledge extraction strategy to obtain virtual power plant data; An operation strategy module, used to determine the virtual power plant operation strategy based on the virtual power plant data; An indicator module is used to determine the control economy evaluation index, the control power supply capacity evaluation index and the control credit index according to the virtual power plant operation strategy; An evaluation module is used to evaluate resource regulation capabilities based on regulation economic evaluation indicators, regulation power supply capacity evaluation indicators, and regulation credit indicators, and obtain regulation data; The control module is used to control the load resources of the virtual power plant according to the control data.

Images