Virtual power plant clearing management system and method based on multi-energy cooperation

Abstract

The invention relates to the field of intelligent management, and provides a virtual power plant clearing management system and method based on multi-energy collaboration, and the method comprises the steps: firstly, enabling each virtual power plant to receive a market adjustment instruction issued by a virtual power plant dispatching center, formulating day-ahead declaration data, and submitting the day-ahead declaration data to the dispatching center; data are integrated in a dispatching center, the integrated data are transmitted to a virtual power plant clearing device to be processed to obtain a day-ahead market clearing result, then a day-ahead pre-dispatching plan is made according to the day-ahead market clearing result, then each virtual power plant receives the plan and distributes the plan to each distributed resource, and when the virtual power plant clearing device receives an emergency regulation and control demand, the virtual power plant clearing device sends the emergency regulation and control demand to the virtual power plant. And if yes, determining an intraday market clearing result and generating an intraday regulation and control instruction, and after receiving the instruction, each virtual power plant allocates the instruction to each distributed resource. Therefore, the virtual power plant resources can be effectively dispatched and managed, and the operation efficiency and economical efficiency of the whole power system are improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent management, and more specifically, to a virtual power plant clearing management system and method based on multi-energy synergy. Background Art

[0002] With the rapid development of renewable energy and the widespread application of distributed energy resources (DERs), the operation mode of traditional power systems faces unprecedented challenges. The traditional centralized power generation model is gradually shifting to a more flexible and decentralized power system, which requires power dispatching and management mechanisms to quickly respond to market changes, optimize resource allocation, and ensure the safe and stable operation of the power grid. As an emerging technology, virtual power plants achieve effective management and dispatch of multiple distributed energy resources, such as wind power, solar energy, and energy storage systems, by aggregating and optimizing these resources, thereby improving the flexibility and reliability of the overall power system.

[0003] Patent CN112780884B proposes a virtual power plant clearing management method, system, device and storage medium, which optimizes the day-ahead and intraday market clearing processes through the collaborative work of the virtual power plant dispatching platform, operator platform and clearing device. The day-ahead process includes receiving market regulation demand, preparing application plans, aggregating data, clearing calculations and forming pre-dispatch plans, while the intraday process performs real-time clearing calculations and generates regulation instructions for emergency regulation needs.

[0004] In the above-mentioned patents, intraday market clearing calculations are performed through specific algorithms (such as optimization algorithms, economic scheduling algorithms, etc.). However, predefined rules and models are often based on historical data and experience settings, and react slowly to emergencies or unforeseen changes. For example, when extreme weather, emergencies, or sudden changes in market demand occur, traditional fixed rules may not be adjusted in time, resulting in delayed scheduling decisions. In addition, parameters and rules are usually fixed and difficult to adjust dynamically to cope with real-time changes, which limits the system's ability to respond quickly to new information, especially in situations where decisions need to be made quickly, such as emergency regulation needs or large market fluctuations.

[0005] Therefore, a virtual power plant clearing management solution based on multi-energy synergy is desired. Summary of the invention

[0006] In response to the shortcomings of the prior art, the present application provides a virtual power plant clearing management system and method based on multi-energy synergy.

[0007] According to one aspect of the present application, a virtual power plant clearing management method based on multi-energy collaboration is provided, including: in response to each virtual power plant receiving a market regulation instruction issued by a virtual power plant dispatching center, formulating day-ahead declaration data and submitting it to the virtual power plant dispatching center; the virtual power plant dispatching center integrates the day-ahead declaration data of each virtual power plant, and transmits the integrated day-ahead declaration data to the virtual power plant clearing device for processing to obtain a day-ahead market clearing result; the day-ahead market clearing result is transmitted to the virtual power plant dispatching center to formulate a day-ahead pre-dispatch plan; each virtual power plant receives the day-ahead pre-dispatch plan and allocates it to each distributed resource; in response to the virtual power plant clearing device receiving an emergency regulation demand, determining the intraday market clearing result and generating an intraday regulation instruction; each virtual power plant receives the intraday regulation instruction and allocates it to each distributed resource; wherein, in response to the virtual power plant clearing device receiving an emergency regulation demand, determining the day-ahead market clearing result The market clearing result includes: collecting virtual power plant market-related information, and simultaneously receiving load power consumption time series data of distributed resources from multiple virtual power plants, wherein the virtual power plant market-related information includes the emergency regulation demand, the market declaration information sealed a day ago, and the day-ahead market clearing result; performing data sorting and load power consumption time series encoding on the load power consumption time series data of the distributed resources of the multiple virtual power plants to obtain a set of virtual power plant operating resource time series state representations; performing time series condition deep enhancement on the set of virtual power plant operating resource time series state representations to obtain a set of virtual power plant operating resource time series state enhanced representations; based on the virtual power plant market-related information, performing feature-guided modulation on the set of virtual power plant operating resource time series state enhanced representations to obtain a set of virtual power plant operating resource state representations guided by market status, and based on the set of virtual power plant operating resource state representations guided by market status, obtaining the intraday market clearing result.

[0008] According to another aspect of the present application, a virtual power plant clearing management system based on multi-energy synergy is provided, which is used to execute the virtual power plant clearing management method based on multi-energy synergy as described above, and includes:

[0009] A data collection module is used to collect information related to the virtual power plant market and simultaneously receive time series data of load power consumption of distributed resources from multiple virtual power plants, wherein the information related to the virtual power plant market includes the emergency regulation demand, the market declaration information sealed the day before, and the market clearing result the day before;

[0010] A virtual power plant operating resource time series data encoding module, used for data sorting and load power consumption time series data of distributed resources of the plurality of virtual power plants to obtain a set of virtual power plant operating resource time series state representations;

[0011] A virtual power plant operating resource timing feature enhancement module, used for performing timing condition depth enhancement on the set of timing state representations of the virtual power plant operating resources to obtain a set of enhanced timing state representations of the virtual power plant operating resources;

[0012] The intraday market clearing result generation module is used to perform feature-guided modulation on the set of enhanced representations of the time series states of the virtual power plant's operating resources based on the virtual power plant market-related information to obtain a set of virtual power plant operating resource state representations guided by the market state, and obtain the intraday market clearing result based on the set of virtual power plant operating resource state representations guided by the market state.

[0013] This application has significant technical effects due to the adoption of the above technical solutions:

[0014] The virtual power plant clearing management system and method based on multi-energy collaboration provided in this application uses artificial intelligence-based data analysis and integration technology to organize and time-series encode the load power consumption time series data of distributed resources of multiple virtual power plants, and then perform time-series enhancement on the time series state representation of the virtual power plant operating resources, and embed the encoding of the market-related information of each virtual power plant, so as to automatically obtain the intraday market clearing results by modulating the enhanced representation of the time series state of each enhanced virtual power plant operating resource based on the characteristics of each embedded code. In this way, the system can quickly respond to emergencies and changes in market demand, so as to quickly adjust the scheduling strategy when extreme weather or emergencies occur, and avoid decision-making delays caused by fixed rules. And the system not only takes into account the current market conditions, but also combines historical load power consumption data, thereby improving the accuracy and reliability of intraday market clearing. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 This is a flow chart of a virtual power plant clearing management method based on multi-energy synergy according to an embodiment of the present application.

[0017] Figure 2 It is a flowchart of step S150 in the virtual power plant clearing management method based on multi-energy synergy according to an embodiment of the present application.

[0018] Figure 3This is a flowchart of step S153 in the virtual power plant clearing management method based on multi-energy synergy according to an embodiment of the present application.

[0019] Figure 4 This is a flowchart of step S154 in the virtual power plant clearing management method based on multi-energy synergy according to an embodiment of the present application.

[0020] Figure 5 This is a system block diagram of a virtual power plant clearing management system based on multi-energy synergy according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0022] Based on the problems in the above-mentioned background technology, the present application provides a virtual power plant clearing management method based on multi-energy synergy. Figure 1 Flow chart of a virtual power plant clearing management method based on multi-energy synergy according to an embodiment of the present application. Figure 1 As shown, according to the embodiment of the present application, the virtual power plant clearing management method based on multi-energy collaboration includes: S110, in response to each virtual power plant receiving a market regulation instruction issued by a virtual power plant dispatching center, formulating day-ahead declaration data and submitting it to the virtual power plant dispatching center; S120, the virtual power plant dispatching center integrates the day-ahead declaration data of each virtual power plant, and transmits the integrated day-ahead declaration data to the virtual power plant clearing device for processing to obtain a day-ahead market clearing result; S130, transmitting the day-ahead market clearing result to the virtual power plant dispatching center to formulate a day-ahead pre-dispatch plan; S140, each virtual power plant receives the day-ahead pre-dispatch plan and allocates it to each distributed resource; S150, in response to the virtual power plant clearing device receiving an emergency regulation demand, determining the intraday market clearing result, and generating an intraday regulation instruction; S160, each virtual power plant receives the intraday regulation instruction and allocates it to each distributed resource.

[0023] In step S110, in response to each virtual power plant receiving the market regulation instruction issued by the virtual power plant dispatching center, the day-ahead declaration data is formulated and submitted to the virtual power plant dispatching center. It should be explained that the virtual power plant is a carrier that organically combines distributed generators, controllable loads and distributed energy storage facilities, and realizes the integration and regulation of various distributed energy sources through supporting regulation technology and communication technology, so as to participate in the power market and power grid operation as a special power plant. In a sense, the virtual power plant can be regarded as an advanced regional power centralized management mode, that is, a management method for multi-energy coordination. Specifically, it can aggregate a variety of distributed energy resources, such as wind energy, solar energy, energy storage systems, etc., to achieve effective management and scheduling of these resources, and improve the flexibility and reliability of the overall power system. The virtual power plant dispatching center is responsible for issuing market regulation instructions, receiving the declaration data submitted by each virtual power plant, aggregating and integrating these data and transmitting them to the virtual power plant clearing device for processing, and formulating pre-dispatching plans based on the clearing results. It is the hub of the entire virtual power plant clearing management process, ensuring information flow and collaborative work in each link. Market regulation instructions specifically refer to the adjustment requirements for the output power and electricity of virtual power plants based on factors such as the supply and demand of the power market and the operating status of the power grid. The day-ahead declaration data includes the declared quantity and price data and revenue calculation of virtual power plant operators and distributed resources. The declared quantity and price data reflect the amount of electricity and the corresponding price that each distributed resource plans to provide or consume in the next day (day-ahead), which helps the market understand the supply and demand of resources and thus make reasonable resource allocation. The revenue calculation is based on the declared quantity and price data and possible market conditions to estimate the expected revenue of virtual power plant operators and distributed resources after participating in market transactions. It should be understood that responding to the market regulation instructions issued by the virtual power plant dispatching center, formulating and submitting day-ahead declaration data is the starting point of the entire virtual power plant clearing management process. In this way, the virtual power plant can fully integrate the information of its own distributed resources according to market demand and dispatching requirements, and reflect the adjustable resource volume and transaction constraints in the declared data. This helps the virtual power plant dispatch center to fully understand the resource status and trading willingness of each virtual power plant, thereby providing an accurate data basis for the subsequent pre-dispatch plan formulation, realizing the optimal allocation of virtual power plant resources, improving the operating efficiency and economy of the entire power system, and ensuring that virtual power plants can effectively participate in electricity market transactions and respond to grid operation needs.

[0024] In step S120, the virtual power plant dispatching center integrates the day-ahead declaration data of each virtual power plant, and transmits the integrated day-ahead declaration data to the virtual power plant clearing device for processing to obtain the day-ahead market clearing result. It should be understood that each virtual power plant has a variety of distributed resources under its jurisdiction, and the resource status (such as the power generation capacity of distributed generators, the regulation capacity of controllable loads, the energy storage level of distributed energy storage facilities, etc.) and transaction intentions (declared quantity and price data and profit calculation) of each virtual power plant are different. In order to fully understand the total amount, distribution and supply or demand of all adjustable resources in the entire virtual power plant system at different price levels, so as to have an overall control of system resources, it is first necessary to integrate the day-ahead declaration data of each virtual power plant. Then, considering that the virtual power plant clearing device has the ability to perform complex calculations according to market rules, it can use a specific algorithm to determine the optimal clearing result based on market regulation needs and integrated declaration data. It can take into account a variety of factors, such as the cost of different resources, the transmission constraints of the power grid, the supply and demand balance of the market, etc., and can accurately calculate the most reasonable resource allocation plan under current market conditions, that is, the day-ahead market clearing result. Therefore, the integrated declaration data needs to be transmitted to the virtual power plant clearing device for processing to obtain the day-ahead market clearing result.

[0025] In step S130, the day-ahead market clearing result is transmitted to the virtual power plant dispatching center to formulate a day-ahead pre-dispatch plan. It should be understood that the virtual power plant dispatching center is responsible for the macro-control and resource integration of the entire virtual power plant system. The day-ahead market clearing result is transmitted to the virtual power plant dispatching center, so that the virtual power plant dispatching center can stand at the system level, comprehensively consider the conditions of each virtual power plant and its distributed resources, and formulate a day-ahead pre-dispatch plan that meets the overall operation requirements of the system. This can ensure the coordinated operation between different virtual power plants and between distributed resources, and avoid system incoordination problems caused by local optimization. Through a reasonable day-ahead pre-dispatch plan, each virtual power plant and its distributed resources can produce or consume according to the optimal quantity-price relationship determined by market clearing, thereby maximizing benefits or minimizing costs. For example, distributed power generation resources can generate more electricity during periods with higher electricity prices to obtain more benefits, controllable loads can increase electricity consumption during periods with lower electricity prices, and energy storage systems can charge during low-price periods and discharge during high-price periods to obtain price difference benefits, thereby improving the economic benefits of virtual power plants as a whole.

[0026] In step S140, each virtual power plant receives the day-ahead pre-dispatch plan and allocates it to each distributed resource. It should be understood that the day-ahead pre-dispatch plan formulated by the virtual power plant dispatch center plans the operation strategy of the entire virtual power plant from the system level, but these plans need to be implemented specifically on each distributed resource before they can be executed. As an entity that directly manages distributed resources, each virtual power plant receives and allocates the day-ahead pre-dispatch plan, which is a key step in converting system-level planning into actual operations, ensuring that each distributed resource clearly knows the tasks of power generation, energy storage or load regulation that it should undertake at different time periods. In this way, each distributed resource can perform power generation, energy storage or load regulation operations at the appropriate time according to market price signals and system operation requirements. In this way, flexible adjustment can be achieved, which is conducive to improving the overall utilization efficiency of various distributed resources and reducing resource waste.

[0027] In step S150, in response to the virtual power plant clearing device receiving an emergency regulation demand, the intraday market clearing result is determined, and an intraday regulation instruction is generated. It should be understood that in the power system, emergencies (such as extreme weather, equipment failure, unexpected power outages, etc.) may have a serious impact on the stability and reliability of the power grid. Therefore, by quickly determining the intraday market clearing result and generating a regulation instruction, the balance of power supply and demand can be adjusted in time to ensure the safe and stable operation of the system.

[0028] Accordingly, in response to the virtual power plant clearing device receiving the emergency regulation demand and determining the intraday market clearing result, the technical concept of the present application is to collect virtual power plant market related information (emergency regulation demand, market declaration information sealed before and market clearing result before), and receive the load power consumption time series data of distributed resources of multiple virtual power plants, and use artificial intelligence-based data analysis and integration technology to sort and time series encode the time series data of each load power consumption, and then enhance the time series state representation of the virtual power plant operating resources, and embed the market related information of each virtual power plant, so as to modulate the enhanced representation of the time series state of each enhanced virtual power plant operating resource based on the characteristics of each embedded code to automatically obtain the intraday market clearing result. In this way, the system can quickly respond to emergencies and changes in market demand, so as to quickly adjust the scheduling strategy when extreme weather or emergencies occur, and avoid decision lags caused by fixed rules. And the system not only takes into account the current market conditions, but also combines historical load power consumption data, thereby improving the accuracy and reliability of intraday market clearing.

[0029] Figure 2 FIG. 1 is a flowchart of step S150 in the virtual power plant clearing management method based on multi-energy synergy according to an embodiment of the present application. Figure 2As shown, the step S150 includes: S151, collecting virtual power plant market-related information, and simultaneously receiving load power consumption time series data of distributed resources from multiple virtual power plants, wherein the virtual power plant market-related information includes the emergency regulation demand, the market declaration information sealed a day ago, and the day-ahead market clearing result; S152, data sorting and load power consumption time series encoding of the load power consumption time series data of the distributed resources of the multiple virtual power plants to obtain a set of virtual power plant operating resource time series state representations; S153, performing time series condition deep enhancement on the set of virtual power plant operating resource time series state representations to obtain a set of virtual power plant operating resource time series state enhanced representations; S154, based on the virtual power plant market-related information, feature-guided modulation on the set of virtual power plant operating resource time series state enhanced representations to obtain a set of virtual power plant operating resource state representations guided by market status, and based on the set of virtual power plant operating resource state representations guided by market status, obtain the intraday market clearing result.

[0030] In step S151, information related to the virtual power plant market is collected, and the load power consumption time series data of distributed resources from multiple virtual power plants are received at the same time, wherein the information related to the virtual power plant market includes the emergency regulation demand, the market declaration information sealed a day ago, and the day-ahead market clearing result. It should be understood that the emergency regulation demand refers to the requirement of the power system for the virtual power plant to quickly adjust the resource output or load regulation in a specific emergency situation (such as power grid failure, sudden major power load changes, etc.); the market declaration information sealed a day ago includes the information on the amount of adjustable resources, transaction intentions and revenue forecasts reported by each virtual power plant and its distributed resources in the day-ahead market stage, which reflects the resource status and market expectations of the virtual power plant in the normal planning cycle, and helps to understand the original resource allocation plan and potential of each virtual power plant; the day-ahead market clearing result is the optimal resource allocation plan calculated by the virtual power plant clearing device in the day-ahead market transaction process according to market rules, the declaration data of each virtual power plant, etc., including the final power generation plan, load regulation arrangement and corresponding transaction price of each virtual power plant. It reflects the resource allocation status after the day-ahead market supply and demand balance. In case of intraday emergency regulation, it can help analyze the gap between the current system resource allocation and emergency demand, and provide a basis for adjustment. The load power consumption time series data of the distributed resources of multiple virtual power plants includes the load power consumption of the distributed resources under the jurisdiction of multiple virtual power plants within a certain time series. It can reflect the real-time operation status and load change trend of each distributed resource. In general, the emergency regulation demand in the virtual power plant market-related information is the key driving factor for determining the direction of intraday market clearing. For example, if the emergency demand is to increase the power generation, then the clearing calculation will be carried out in the direction of prompting each virtual power plant to increase the input of power generation resources. The market declaration information sealed the day before and the day-ahead market clearing results provide the initial resource allocation status and reference standards for the calculation. For example, according to the comparison between the power generation plan the day before and the current emergency demand, the amplitude and scope of the adjustment can be determined. The load and electricity consumption time series data of distributed resources of multiple virtual power plants is the basis for the real-time status calculation. By analyzing the load change trends, real-time output of power generation resources and other conditions in these data, combined with emergency control needs, a specific algorithm is used to calculate the optimal resource reallocation plan that meets emergency control needs under the current system state, that is, the intraday market clearing result.

[0031] In step S152, the load power consumption time series data of the distributed resources of the multiple virtual power plants are sorted and the load power consumption time series is encoded to obtain a set of virtual power plant operating resource time series state representations. Specifically, in an embodiment of the present application, the step S152 includes: sorting the load power consumption time series data of the distributed resources of the multiple virtual power plants according to the time dimension to obtain a data set of virtual power plant distributed resource load power consumption time series input vectors; passing each virtual power plant distributed resource load power consumption time series input vector in the data set of the virtual power plant distributed resource load power consumption time series input vector through a load power consumption time series encoder based on a forward LSTM model to obtain a set of virtual power plant operating resource time series state representation vectors as the set of virtual power plant operating resource time series state representations.

[0032] It should be understood that the load power consumption time series data has time series characteristic information in the time dimension. In order to ensure that all collected data are aligned in the time dimension and adopt a consistent format and structure to better capture and mine the time series characteristics, in the technical solution of the present application, the load power consumption time series data of the distributed resources of the multiple virtual power plants are sorted according to the time dimension to obtain a data set of the virtual power plant distributed resource load power consumption time series input vector, which can ensure data consistency and provide data support for subsequent time series editing.

[0033] Then, considering that the distributed resource load power consumption time series input vectors of each virtual power plant have different time series correlations in different time intervals, such as the continuous increase in power consumption in a short period of time or the occurrence of peaks, etc. Therefore, in order to effectively extract important features and patterns that change over time from historical load power consumption data, in the technical solution of the present application, each virtual power plant distributed resource load power consumption time series input vector in the data set of the virtual power plant distributed resource load power consumption time series input vector is respectively passed through a load power consumption time series encoder based on a forward LSTM model to utilize the characteristics of the forward LSTM model that is good at processing and learning long-term dependencies in time series data to capture and extract the time series information of power consumption in different local time periods, and obtain a set of virtual power plant operating resource time series state representation vectors.

[0034] In step S153, the set of virtual power plant operating resource time series state representations is subjected to time series condition depth enhancement to obtain a set of virtual power plant operating resource time series state enhanced representations. Specifically, Figure 3 FIG. 1 is a flowchart of step S153 in the virtual power plant clearing management method based on multi-energy synergy according to an embodiment of the present application. Figure 3As shown, the step S153 includes: S1531, performing field mapping on each virtual power plant operating resource timing state characterization vector in the set of virtual power plant operating resource timing state characterization vectors to obtain a set of entry virtual power plant operating resource coding feature vectors; S1532, based on the set of entry virtual power plant operating resource coding feature vectors, performing timing feature enhancement on the set of virtual power plant operating resource timing state characterization vectors to obtain a set of virtual power plant operating resource timing state enhanced characterization vectors as the set of virtual power plant operating resource timing state enhanced characterizations.

[0035] It should be understood that there are different characteristics between different virtual power plants, that is, there are significant key information to varying degrees in the timing state representation of the operating resources of each virtual power plant, and each representation information occupies a different proportion in the power consumption information of the entire power plant. Therefore, in order to highlight the key deep characteristics between different virtual power plants, make the unique mode and behavior of each distributed resource more obvious, and provide better data support for the subsequent allocation of resources, the technical solution of this application introduces a timing condition deep enhancement mechanism to process the set of timing state representations of the operating resources of the virtual power plant to obtain a set of enhanced timing state representations of the operating resources of the virtual power plant. That is, this enhancement can not only make the characteristics of each virtual power plant more obvious, but also adjust the feature weight according to its relative importance in the overall power consumption information, thereby providing more accurate data support for subsequent resource allocation decisions.

[0036] Specifically, first, field mapping is performed on each virtual power plant operating resource time series state representation vector in the set of virtual power plant operating resource time series state representation vectors to obtain a set of virtual power plant operating resource encoding feature vectors. The above process can be expressed as:

[0037] h i =W 1 I i W 2

[0038] Among them, I i is the i-th virtual power plant operating resource timing state representation vector in the set of the virtual power plant operating resource timing state representation vectors, W 1 and W 2 are the first weight matrix and the second weight matrix respectively, h i isI i The corresponding entry virtual power plant operation resource encoding feature vector.

[0039] That is, by performing a field mapping operation on each virtual power plant operating resource time series state representation vector in the set of virtual power plant operating resource time series state representation vectors, these virtual power plant operating resource time series state characteristics can be converted into a form that is more suitable for the current task, that is, a set of virtual power plant operating resource encoding feature vectors. The essence of this mapping operation is a transformation of the feature space, which enables the original features to better reflect the internal relationship between the data, and thus be more easily captured by the algorithm.

[0040] Specifically, in an embodiment of the present application, the step S1532 includes: calculating the field depth factor of each entry virtual power plant operation resource coding feature vector in the set of entry virtual power plant operation resource coding feature vectors to obtain a set of virtual power plant operation resource field depth factors; based on the set of virtual power plant operation resource field depth factors, calculating the virtual power plant operation resource field essential feature vector of the set of entry virtual power plant operation resource coding feature vectors; using the virtual power plant operation resource field essential feature vector as a conditional feature vector, respectively, distinguishably enhancing each virtual power plant operation resource timing state representation vector in the set of virtual power plant operation resource timing state representation vectors to obtain a set of virtual power plant operation resource timing state enhanced representation vectors.

[0041] Next, the field depth factor of each entry virtual power plant operation resource encoding feature vector in the set of entry virtual power plant operation resource encoding feature vectors is calculated to obtain a set of virtual power plant operation resource field depth factors, and based on the set of virtual power plant operation resource field depth factors, the virtual power plant operation resource field essential feature vector of the set of entry virtual power plant operation resource encoding feature vectors is calculated. The above process can be expressed as:

[0042]

[0043] Among them, h i isI i The corresponding entry virtual power plant operation resource encoding feature vector, ||·|| 2 To calculate the square of the vector length, f(h i ) is used to calculate h i The depth of field factor, e i h i The corresponding virtual power plant operation resource field depth factor, exp(·) represents the exponential function value with the natural constant e as the base, a i For I i The corresponding standardized virtual power plant operation resource field depth factor, v c It is the essential characteristic vector of the virtual power plant operating resource field.

[0044] That is, for each incoming virtual power plant operation resource coding feature vector after field mapping, the corresponding field depth factor is further calculated. The field depth factor can be understood as an indicator to measure the relative importance or complexity of the feature point in the area to which it belongs. By calculating such factors one by one for all the field virtual power plant operation resource coding feature vectors in the entire set, a set of virtual power plant operation resource field depth factors can be obtained. Then, using the above-obtained set of virtual power plant operation resource field depth factors as the basis, the field effect normalization mechanism is used to extract a feature distribution that comprehensively reflects the core of the entire batch of incoming virtual power plant operation resource coding feature vectors, that is, the obtained virtual power plant operation resource field essential feature vector.

[0045] More specifically, in an embodiment of the present application, the virtual power plant operating resource field essential characteristic vector is used as a conditional characteristic vector, and each virtual power plant operating resource timing state representation vector in the set of the virtual power plant operating resource timing state representation vector is distinguishably enhanced to obtain a set of virtual power plant operating resource timing state enhanced representation vectors, including: using a sigmoid function to fully connect encode the virtual power plant operating resource field essential characteristic vector and each virtual power plant operating resource timing state representation vector to obtain a set of virtual power plant operating resource timing state global and local interaction vectors; the set of virtual power plant operating resource timing state global and local interaction vectors and the corresponding virtual power plant operating resource timing state representation vector in the set of virtual power plant operating resource timing state representation vectors are multiplied by position point to obtain a set of virtual power plant operating resource timing state enhanced representation vectors. The above process can be expressed as:

[0046] V si =I i ⊙[Sigmoid(M m v c +H m I i +b m )]

[0047] Among them, I i is the i-th virtual power plant operating resource timing state representation vector in the set of the virtual power plant operating resource timing state representation vectors, v c is the essential characteristic vector of the virtual power plant operation resource field, M m is the virtual power plant operation resource essential weight matrix, H m is the timing weight matrix of local virtual power plant operation resources, b m is the bias vector, sigmoid(·) is the sigmoid function, ⊙ is the position point multiplication, V si Yesi The enhanced representation vector of the temporal state of the virtual power plant operating resources after semantic enhancement.

[0048] That is, using the essential characteristic vector of the virtual power plant operating resource field as additional conditional information, each virtual power plant operating resource timing state representation vector in the set of virtual power plant operating resource timing state representation vectors is distinguishably enhanced. This facilitates flexible adjustment of parameter configuration according to the needs of actual application scenarios, and can specifically amplify important features of certain aspects of the timing state of virtual power plant operating resources, while suppressing other less important parts, so as to optimize the expression of the characteristics of the timing state of virtual power plant operating resources. The enhanced representation vectors of the timing state of each virtual power plant operating resource obtained are more prominent and easy to distinguish, and can more effectively express the deep information mined from the original data, thereby significantly improving the quality of feature representation.

[0049] In step S154, based on the virtual power plant market related information, the set of enhanced representations of the time series state of the virtual power plant operating resources is feature-guided modulated to obtain a set of virtual power plant operating resource state representations guided by market conditions, and based on the set of virtual power plant operating resource state representations guided by market conditions, the intraday market clearing result is obtained. Specifically, Figure 4 FIG. 1 is a flowchart of step S154 in the virtual power plant clearing management method based on multi-energy synergy according to an embodiment of the present application. Figure 4 As shown, the step S154 includes: S1541, respectively embedding and encoding the emergency regulation demand, the market declaration information sealed a day ago and the day-ahead market clearing result in the virtual power plant market related information to obtain the emergency regulation demand semantic embedded coding feature vector, the day-ahead market declaration information semantic embedded coding feature vector and the day-ahead market clearing result semantic embedded coding feature vector; S1542, based on the emergency regulation demand semantic embedded coding feature vector, the day-ahead market declaration information semantic embedded coding feature vector and the day-ahead market clearing result semantic embedded coding feature vector, modulating each virtual power plant operating resource timing state enhanced representation vector in the set of virtual power plant operating resource timing state enhanced representation vector to obtain a set of virtual power plant operating resource state representation vectors under market state guidance as the set of virtual power plant operating resource state representation under market state guidance; S1543, based on the set of virtual power plant operating resource state representation under market state guidance, obtaining the intraday market clearing result.

[0050] In step S1541, the emergency regulation demand, the market declaration information sealed the day before, and the day-ahead market clearing result in the virtual power plant market-related information are embedded and coded respectively to obtain the semantic embedded coding feature vector of the emergency regulation demand, the semantic embedded coding feature vector of the day-ahead market declaration information, and the semantic embedded coding feature vector of the day-ahead market clearing result. Accordingly, considering that each piece of information in the virtual power plant market-related information contains deep and implicit semantic information, in order to understand and analyze the key semantic information contained in each piece of information, in the technical solution of the present application, the emergency regulation demand, the market declaration information sealed the day before, and the day-ahead market clearing result in the virtual power plant market-related information are embedded and coded respectively to capture the core meaning and contextual relationship of the data, and obtain the semantic embedded coding feature vector of the emergency regulation demand, the semantic embedded coding feature vector of the day-ahead market declaration information, and the semantic embedded coding feature vector of the day-ahead market clearing result. In particular, in a specific embodiment of the present application, a Bert model-based encoder including a word embedding layer can be used to embed code the emergency regulation demand in the virtual power plant market-related information after word segmentation preprocessing, the day-ahead market declaration information sealed after word segmentation preprocessing, and the day-ahead market clearing results after word segmentation preprocessing, respectively, to obtain the emergency regulation demand semantic embedding coding feature vector, the day-ahead market declaration information semantic embedding coding feature vector, and the day-ahead market clearing result semantic embedding coding feature vector.

[0051] In step S1542, based on the semantic embedding coding feature vector of the emergency regulation demand, the semantic embedding coding feature vector of the day-ahead market declaration information, and the semantic embedding coding feature vector of the day-ahead market clearing result, each virtual power plant operating resource time series state enhanced representation vector in the set of virtual power plant operating resource time series state enhanced representation vectors is modulated to obtain a set of virtual power plant operating resource state representation vectors guided by market state as the set of virtual power plant operating resource state representations guided by market state. It should be understood that market conditions (such as the semantic embedding coding feature of the emergency regulation demand) and historical data (such as the semantic embedding coding feature of the day-ahead market declaration information and the semantic embedding coding feature of the day-ahead market clearing result) directly affect the operation strategy of the virtual power plant, providing external guidance and reference for scheduling decisions, while the enhanced representation of the time series state of each virtual power plant operating resource reflects the actual operating state of the internal resources of the virtual power plant, providing basic data for scheduling decisions. Therefore, in order to better reflect the best operating mode of each distributed resource under the current market conditions and realize the optimal configuration of resources, in the technical solution of the present application, each virtual power plant operating resource timing state enhanced representation vector in the set of the virtual power plant operating resource timing state enhanced representation vector is modulated based on the semantic embedded coding feature vector of the emergency regulation demand, the semantic embedded coding feature vector of the day-ahead market declaration information and the semantic embedded coding feature vector of the day-ahead market clearing result. In this way, it can be ensured that the generated state representation vector is highly correlated with the current market state, thereby providing a basis for formulating a more accurate scheduling plan. In particular, in the embodiment of the present application, the implementation process of step S1542 can be expressed by the formula:

[0052]

[0053] V mi =w s ·V si

[0054] Among them, V a 、V b and V d are the semantic embedding encoding feature vectors of the day-ahead market clearing result, the day-ahead market declaration information, and the semantic embedding encoding feature vector of the emergency regulation demand, respectively. exp(·) represents the exponential function value with the natural constant e as the base. softmax(·) is the softmax function. w s is the operating resource timing weight vector, V si is the i-th virtual power plant operating resource timing state enhancement representation vector in the set of virtual power plant operating resource timing state enhancement representation vectors, V miIt is the i-th virtual power plant operating resource state representation vector under market state guidance in the set of virtual power plant operating resource state representation vectors under market state guidance.

[0055] In step S1543, based on the set of virtual power plant operating resource state representations under the market state guidance, the intraday market clearing result is obtained. Specifically, in the embodiment of the present application, the step S1543 includes: passing each virtual power plant operating resource state representation vector under the market state guidance in the set of virtual power plant operating resource state representation vectors under the market state guidance through a decoder-based demand resource allocator to obtain a demand resource allocation ratio value of each virtual power plant; multiplying the emergency regulation demand by the demand resource allocation ratio value of each virtual power plant and accumulating them to obtain the intraday market clearing result. More specifically, in an embodiment of the present application, each market-state guided virtual power plant operating resource state representation vector in the set of market-state guided virtual power plant operating resource state representation vectors is respectively passed through a decoder-based demand resource allocator to obtain a demand resource allocation ratio value of each virtual power plant, including: using the fully connected layer of the decoder to decode and regress each market-state guided virtual power plant operating resource state representation vector in the set of market-state guided virtual power plant operating resource state representation vectors to obtain a demand resource allocation ratio value of each virtual power plant.

[0056] It should be understood that although the virtual power plant operating resource state representation vector guided by the market state contains rich information about the virtual power plant operating state and market conditions, this information is an abstract representation after a series of processing. In order to convert this information into a specific proportional value that can be directly used for resource allocation decisions, a decoder is required for conversion. The decoder is a neural network model that can decode the input vector data according to the learned pattern through a large amount of data training, and convert it into a specific value that can be used to guide the resource allocation of the virtual power plant. The demand resource allocation ratio value of each virtual power plant obtained by decoding refers to the resource allocation ratio obtained by each virtual power plant under specific emergency regulation requirements and market conditions, which guides the resource allocation between virtual power plants. Finally, the emergency regulation demand is multiplied by the demand resource allocation ratio value of each virtual power plant and then accumulated to obtain the intraday market clearing result. In this way, the specific resource allocation amount that each virtual power plant should bear under emergency regulation can be accurately calculated. Moreover, the accumulated intraday market clearing results can take into account the urgency of market demand and the rationality of virtual power plant resource allocation, ensuring that resource allocation decisions not only meet the needs of current emergency regulation, but also achieve fair and effective allocation among different virtual power plants.

[0057] In particular, in one example, since each virtual power plant operating resource timing state enhanced representation vector in the set of virtual power plant operating resource timing state enhanced representation vectors respectively represents the full time domain sequence descriptive distinction enhanced timing correlation characteristics of the load power consumption of the distributed resources of each virtual power plant, after being modulated by the embedded coding features of the virtual power plant market-related information, it is expected to improve the overall semantically consistent aggregation expression effect of the set of virtual power plant operating resource state representation vectors guided by the market status.

[0058] Based on this, the present application optimizes the set of virtual power plant operating resource state representation vectors guided by the market state when each of the virtual power plant operating resource state representation vectors guided by the market state in the set of virtual power plant operating resource state representation vectors guided by the market state is passed through a decoder-based demand resource allocator to obtain the demand resource allocation amount of the multiple virtual power plants, including:

[0059] Performing feature clustering on all feature values ​​of the set of virtual power plant operating resource state characterization vectors guided by the market state to obtain a feature set within the virtual power plant operating resource state class guided by the market state The out-of-class feature set of virtual power plant operating resource status under the guidance of market status

[0060] Calculate the ratio λ=k / n of the number of feature values ​​in the feature set within the virtual power plant operating resource state class under the market state guidance to the number of feature values ​​of all features of the set of virtual power plant operating resource characterization vectors under the market state guidance;

[0061] Calculate the ratio of the λth power of the sum of the absolute values ​​of all eigenvalues ​​in the feature set within the virtual power plant operating resource state class under the market state guidance to the λth power of the sum of the absolute values ​​of all eigenvalues ​​in the set of the virtual power plant operating resource state characterization vector under the market state guidance to obtain the first virtual power plant operating resource state weight under the market state guidance;

[0062] Calculate the ratio of the λ / 2 power of the sum of squares of all eigenvalues ​​in the feature set within the virtual power plant operating resource state class under the market state guidance to the λ / 2 power of the sum of squares of all eigenvalues ​​in the set of the virtual power plant operating resource state characterization vector under the market state guidance to obtain the second virtual power plant operating resource state weight under the market state guidance;

[0063] That is, the virtual power plant operating resource state weight w under the guidance of the first market state 1 and the virtual power plant operation resource state weight w under the second market state guidance 2 It is expressed as:

[0064]

[0065] Among them, f 1i represents each feature value in the feature set of the virtual power plant operating resource state class under the guidance of the market state, f p represents each eigenvalue in the set of virtual power plant operating resource characterization vectors under the market status, k represents the number of eigenvalues ​​in the feature set within the virtual power plant operating resource state class under the market status, n represents the number of eigenvalues ​​of all features in the set of virtual power plant operating resource characterization vectors under the market status, λ represents the ratio between k and n, w 1 represents the weight of the virtual power plant operating resource state under the guidance of the first market state, w 2 represents the weight of the virtual power plant operating resource state under the guidance of the second market state;

[0066] Then, for each feature value in the feature set within the virtual power plant operating resource state class guided by the market state, the product of the feature value and the weight of the virtual power plant operating resource state guided by the second market state is calculated, and then the weight of the virtual power plant operating resource state guided by the first market state is added to obtain the optimized feature value; this process can be expressed as:

[0067] f′ 1i =w 2 ×f 1i +w 1

[0068] Among them, w 1 represents the weight of the virtual power plant operating resource state under the guidance of the first market state, w 2 represents the weight of the virtual power plant operating resource state under the guidance of the second market state, f 1i represents each feature value in the feature set of the virtual power plant operating resource state class under the guidance of the market state, f' 1i Indicates the optimized feature values ​​of each feature value in the feature set within the virtual power plant operating resource state class under the guidance of the market state;

[0069] For each feature value in the out-of-class feature set of the virtual power plant operating resource state under the market state guidance, the product of the feature value and the weight of the virtual power plant operating resource state under the first market state guidance is calculated to obtain the optimized feature value; this process can be expressed as:

[0070] f 2j '=w 1 ×f 2j

[0071] Among them, w 1represents the weight of the virtual power plant operating resource state under the guidance of the first market state, f 2j represents each feature value in the out-of-class feature set of the virtual power plant operating resource state under the guidance of the market state, f 2j ' represents the optimized feature values ​​of each feature value in the feature set outside the virtual power plant operating resource state class under the guidance of the market state;

[0072] Finally, the optimized feature values ​​of the feature set within the virtual power plant operating resource state class under the guidance of the market state and the feature set outside the virtual power plant operating resource state class under the guidance of the market state are used to form a set of optimized characterization vectors of the virtual power plant operating resource state under the guidance of the market state.

[0073] That is, while the set of virtual power plant operating resource state characterization vectors guided by the market status is feature clustered, the interactive description of the key feature information of the set of virtual power plant operating resource state characterization vectors guided by the market status in the clustering process is carried out, and the geometric equivariant topology of the features is constructed by the low-rank harmonic modulation of the clustering features of the set of virtual power plant operating resource state characterization vectors guided by the market status and the equivariance of the features as a whole, so as to obtain the graphical distribution translation and rotation symmetry of the clustering features of the set of virtual power plant operating resource state characterization vectors guided by the market status relative to the features as a whole, so as to introduce geometric message passing into the feature expression of the set of virtual power plant operating resource state characterization vectors guided by the market status through the manipulation of irreducible low-rank coefficients to achieve the clustering mapping symmetry of the set of virtual power plant operating resource state characterization vectors guided by the market status, improve the consistency of the clustering-based feature representation of the set of virtual power plant operating resource state characterization vectors guided by the market status, and thus improve the consistency of the demand resource allocation amounts of the multiple virtual power plants obtained by the decoder-based demand resource allocator of the virtual power plant operating resource state characterization vectors guided by each market status.

[0074] In summary, step S150 is explained clearly, which uses artificial intelligence-based data analysis and integration technology to organize and time-series encode the load power consumption time series data of distributed resources of multiple virtual power plants, and then perform time-series enhancement on the time series state representation of the virtual power plant operating resources, and embed the market-related information of each virtual power plant, so as to automatically obtain the intraday market clearing result by modulating the enhanced representation of the time series state of each enhanced virtual power plant operating resource based on the characteristics of each embedded code. In this way, the system can quickly respond to emergencies and changes in market demand, so as to quickly adjust the scheduling strategy when extreme weather or emergencies occur, and avoid decision-making delays caused by fixed rules. And the system not only takes into account the current market conditions, but also combines historical load power consumption data, thereby improving the accuracy and reliability of intraday market clearing.

[0075] In step S160, each virtual power plant receives the intraday control instruction and distributes it to each distributed resource. It should be understood that each virtual power plant receives the intraday control instruction, and can accurately transmit the control decision made by the virtual power plant clearing device to the specific execution level, that is, each distributed resource. Each distributed resource adjusts its own operating parameters or working status according to the received instructions, so as to achieve a rapid response to the market's emergency control needs. For example, in an emergency situation where it is necessary to increase the power generation capacity, the distributed power source can increase the output according to the instruction, and the energy storage system can perform reasonable charging and discharging according to the instruction, so that the power generation and regulation capabilities of the entire virtual power plant system can quickly adapt to market changes.

[0076] Through virtual power plants, multi-energy coordination can be achieved, mainly including the following points: a. Virtual power plants have strong resource integration capabilities. They integrate distributed generators (covering traditional energy and new energy), controllable loads and distributed energy storage facilities into a unified resource pool through control and communication technology, which is convenient for centralized management and scheduling. b. The intelligent control technology used by virtual power plants can formulate optimal strategies based on energy characteristics, market demand and adjustable resource quantities. For example, solar energy is given priority when there is sufficient sunshine during the day, and wind power or natural gas power generation is flexibly switched at night or on cloudy days. c. The efficient communication network ensures real-time information interaction between various resource entities within the virtual power plant and the external power grid, so that the virtual power plant can quickly adjust energy output according to the state of the power grid. d. The drive of the market mechanism also prompts virtual power plants to optimize resource allocation according to price signals. In general, the architecture model of virtual power plants can improve energy utilization efficiency and make full use of advantageous energy in various time periods; it can smooth the volatility of new energy and ensure the stability of the power grid; it can also optimize cost-effectiveness and enhance the adaptability of the power grid, thereby realizing the complementary and coordinated dispatch of traditional energy and new energy and the optimized operation of the power grid.

[0077] In summary, the virtual power plant clearing management method based on multi-energy collaboration based on the embodiment of the present application is explained. First, after each virtual power plant receives the market regulation instruction issued by the virtual power plant dispatching center, it formulates the day-ahead declaration data and submits it to the dispatching center. The dispatching center integrates the data and transmits the integrated data to the virtual power plant clearing device for processing to obtain the day-ahead market clearing result. Then, a day-ahead pre-dispatch plan is formulated based on this. Subsequently, each virtual power plant receives the plan and allocates it to each distributed resource. When the virtual power plant clearing device receives an emergency regulation demand, it determines the intraday market clearing result and generates an intraday regulation instruction. After receiving this instruction, each virtual power plant allocates it to each distributed resource. In this way, the effective dispatching and management of virtual power plant resources can be achieved, and the operating efficiency and economy of the entire power system can be improved.

[0078] In an embodiment of the present application, a virtual power plant clearing management system based on multi-energy synergy is also provided, which is used to execute the virtual power plant clearing management method based on multi-energy synergy as described above. Figure 5 : is a system block diagram of a virtual power plant clearing management system based on multi-energy synergy according to an embodiment of the present application. Figure 5 As shown, according to the embodiment of the present application, the virtual power plant clearing management system 100 based on multi-energy collaboration includes: a data acquisition module 110, which is used to collect virtual power plant market related information, and at the same time receive load power consumption time series data of distributed resources from multiple virtual power plants, wherein the virtual power plant market related information includes the emergency regulation demand, the market declaration information sealed a day ago, and the market clearing result a day ago; a virtual power plant operating resource time series data encoding module 120, which is used to sort the load power consumption time series data of the distributed resources of the multiple virtual power plants and load power consumption time series encoding to obtain the virtual power plant operating resource time series data; A set of sequence state representations; a virtual power plant operating resource timing feature enhancement module 130, used to perform timing condition deep enhancement on the set of virtual power plant operating resource timing state representations to obtain a set of virtual power plant operating resource timing state enhanced representations; an intraday market clearing result generation module 140, used to perform feature-guided modulation on the set of virtual power plant operating resource timing state enhanced representations based on the virtual power plant market-related information to obtain a set of virtual power plant operating resource state representations guided by market conditions, and obtain the intraday market clearing result based on the set of virtual power plant operating resource state representations guided by market conditions.

[0079] Here, those skilled in the art can understand that the specific functions and operations of each unit and module in the virtual power plant clearing management system 100 based on multi-energy synergy have been referred to above. Figures 1 to 4 It has been introduced in detail in the description of the virtual power plant clearing management method based on multi-energy synergy, and therefore, its repeated description will be omitted.

[0080] In summary, the virtual power plant clearing management system 100 based on multi-energy collaboration based on the embodiment of the present application is explained. First, after each virtual power plant receives the market regulation instruction issued by the virtual power plant dispatching center, it formulates the day-ahead declaration data and submits it to the dispatching center. The dispatching center integrates the data and transmits the integrated data to the virtual power plant clearing device for processing to obtain the day-ahead market clearing result. Then, a day-ahead pre-dispatch plan is formulated based on this. Subsequently, each virtual power plant receives the plan and allocates it to each distributed resource. When the virtual power plant clearing device receives an emergency regulation demand, it determines the intraday market clearing result and generates an intraday regulation instruction. After receiving this instruction, each virtual power plant allocates it to each distributed resource. In this way, the effective dispatching and management of virtual power plant resources can be achieved, and the operating efficiency and economy of the entire power system can be improved.

Claims

1. A virtual power plant clearing management method based on multi-energy synergy, comprising: In response to each virtual power plant receiving a market regulation instruction issued by a virtual power plant dispatch center, formulating day-ahead declaration data and submitting it to the virtual power plant dispatch center; The virtual power plant dispatch center integrates the day-ahead declaration data of each virtual power plant, and transmits the integrated day-ahead declaration data to the virtual power plant clearing device for processing to obtain the day-ahead market clearing result; transmits the day-ahead market clearing result to the virtual power plant dispatch center to formulate a day-ahead pre-dispatch plan; each virtual power plant receives the day-ahead pre-dispatch plan and allocates it to each distributed resource; In response to the virtual power plant clearing device receiving an emergency regulation demand, determining an intraday market clearing result and generating an intraday regulation instruction; each virtual power plant receives the intraday regulation instruction and allocates it to each distributed resource; It is characterized in that, in response to the virtual power plant clearing device receiving an emergency regulation demand, determining the intraday market clearing result, includes: collecting virtual power plant market related information, and simultaneously receiving load power consumption time series data of distributed resources from multiple virtual power plants, wherein the virtual power plant market related information includes the emergency regulation demand, the market declaration information sealed a day ago, and the day-ahead market clearing result; performing data sorting and load power consumption time series encoding on the load power consumption time series data of the distributed resources of the multiple virtual power plants to obtain a set of virtual power plant operating resource time series state representations; performing time series condition deep enhancement on the set of virtual power plant operating resource time series state representations to obtain a set of virtual power plant operating resource time series state enhanced representations; based on the virtual power plant market related information, performing feature-guided modulation on the set of virtual power plant operating resource time series state enhanced representations to obtain a set of virtual power plant operating resource state representations guided by market status, and based on the set of virtual power plant operating resource state representations guided by market status, obtaining the intraday market clearing result.

2. The virtual power plant clearing management method based on multi-energy synergy according to claim 1 is characterized in that: The load power consumption time series data of the distributed resources of the plurality of virtual power plants are sorted and the load power consumption time series is encoded to obtain a set of virtual power plant operation resource time series state representations, including: Arrange the load power consumption time series data of the distributed resources of the multiple virtual power plants according to the time dimension to obtain a data set of the virtual power plant distributed resource load power consumption time series input vector; Each virtual power plant distributed resource load power consumption timing input vector in the data set of the virtual power plant distributed resource load power consumption timing input vector is respectively passed through the load power consumption timing encoder based on the forward LSTM model to obtain a set of virtual power plant operating resource timing state representation vectors as the set of virtual power plant operating resource timing state representations.

3. The virtual power plant clearing management method based on multi-energy synergy according to claim 2 is characterized in that: The set of the virtual power plant operating resource timing state representations is deeply enhanced under timing conditions to obtain a set of enhanced timing state representations of the virtual power plant operating resources, including: Performing field mapping on each virtual power plant operating resource time series state representation vector in the set of virtual power plant operating resource time series state representation vectors to obtain a set of incoming virtual power plant operating resource encoding feature vectors; Based on the set of encoding feature vectors of the entering virtual power plant operating resources, the set of timing state characterization vectors of the virtual power plant operating resources is subjected to timing feature enhancement to obtain a set of enhanced timing state characterization vectors of the virtual power plant operating resources as the set of enhanced timing state characterizations of the virtual power plant operating resources.

4. The virtual power plant clearing management method based on multi-energy synergy according to claim 3 is characterized in that: Based on the set of encoding feature vectors of the virtual power plant operating resources entering the site, the set of time series state representation vectors of the virtual power plant operating resources is subjected to time series feature enhancement to obtain a set of enhanced time series state representation vectors of the virtual power plant operating resources, including: Calculating the field depth factor of each entry virtual power plant operation resource encoding feature vector in the set of entry virtual power plant operation resource encoding feature vectors to obtain a set of virtual power plant operation resource field depth factors; Based on the set of virtual power plant operation resource field depth factors, calculating the virtual power plant operation resource field essential feature vector of the set of the incoming virtual power plant operation resource encoding feature vectors; Taking the essential characteristic vector of the virtual power plant operating resource field as the conditional characteristic vector, each virtual power plant operating resource timing state representation vector in the set of virtual power plant operating resource timing state representation vectors is distinguishably enhanced to obtain a set of virtual power plant operating resource timing state enhanced representation vectors.

5. The virtual power plant clearing management method based on multi-energy synergy according to claim 4 is characterized in that: Taking the virtual power plant operation resource field essential feature vector as the conditional feature vector, each virtual power plant operation resource timing state representation vector in the set of the virtual power plant operation resource timing state representation vector is distinguishably enhanced to obtain a set of virtual power plant operation resource timing state enhanced representation vectors, including: Using the si gmo id function, the virtual power plant operating resource field essential feature vector and the respective virtual power plant operating resource timing state characterization vectors are fully connected and encoded to obtain a set of global and local interaction vectors of the virtual power plant operating resource timing state; The set of global and local interaction vectors of the virtual power plant operating resource timing states and the set of characterization vectors of the virtual power plant operating resource timing states corresponding to the global and local interaction vectors of the virtual power plant operating resource timing states and the characterization vectors of the virtual power plant operating resource timing states are multiplied point by point to obtain the set of enhanced characterization vectors of the virtual power plant operating resource timing states.

6. The virtual power plant clearing management method based on multi-energy synergy according to claim 5 is characterized in that: Based on the virtual power plant market related information, feature-guided modulation is performed on the set of enhanced representations of the time series state of the virtual power plant operating resources to obtain a set of virtual power plant operating resource state representations guided by market state, and based on the set of virtual power plant operating resource state representations guided by market state, the intraday market clearing result is obtained, including: The emergency regulation demand, the market declaration information sealed a day ago, and the day-ahead market clearing result in the virtual power plant market related information are respectively embedded and coded to obtain the emergency regulation demand semantic embedded coding feature vector, the day-ahead market declaration information semantic embedded coding feature vector, and the day-ahead market clearing result semantic embedded coding feature vector; Based on the semantic embedded coding feature vector of the emergency regulation demand, the semantic embedded coding feature vector of the day-ahead market declaration information and the semantic embedded coding feature vector of the day-ahead market clearing result, each virtual power plant operating resource time series state enhanced representation vector in the set of virtual power plant operating resource time series state enhanced representation vectors is modulated to obtain a set of virtual power plant operating resource state representation vectors guided by market state as the set of virtual power plant operating resource state representations guided by market state; Based on the set of representations of the virtual power plant operating resource states guided by the market state, the intraday market clearing result is obtained.

7. The virtual power plant clearing management method based on multi-energy synergy according to claim 6 is characterized in that: Based on the set of virtual power plant operating resource state representations guided by the market state, the intraday market clearing result is obtained, including: Each of the virtual power plant operating resource state representation vectors guided by the market state in the set of virtual power plant operating resource state representation vectors guided by the market state is passed through a demand resource allocator based on a decoder to obtain a demand resource allocation ratio value of each virtual power plant; The emergency regulation demand is multiplied by the demand resource allocation ratio of each virtual power plant and then accumulated to obtain the intraday market clearing result.

8. The virtual power plant clearing management method based on multi-energy synergy according to claim 7 is characterized in that: Each of the market-state guided virtual power plant operating resource state representation vectors in the set of market-state guided virtual power plant operating resource state representation vectors is passed through a decoder-based demand resource allocator to obtain a demand resource allocation ratio value for each virtual power plant, including: using the fully connected layer of the decoder to decode and regress each of the market-state guided virtual power plant operating resource state representation vectors in the set of market-state guided virtual power plant operating resource state representation vectors to obtain a demand resource allocation ratio value for each virtual power plant.

9. A virtual power plant clearing management system based on multi-energy synergy, used to execute the virtual power plant clearing management method based on multi-energy synergy as claimed in claim 1, characterized in that: The virtual power plant clearing management system based on multi-energy synergy includes: A data collection module is used to collect information related to the virtual power plant market and simultaneously receive time series data of load power consumption of distributed resources from multiple virtual power plants, wherein the information related to the virtual power plant market includes the emergency regulation demand, the market declaration information sealed the day before, and the market clearing result the day before; A virtual power plant operating resource time series data encoding module, used for data sorting and load power consumption time series data of distributed resources of the plurality of virtual power plants to obtain a set of virtual power plant operating resource time series state representations; A virtual power plant operating resource timing feature enhancement module, used for performing timing condition depth enhancement on the set of timing state representations of the virtual power plant operating resources to obtain a set of enhanced timing state representations of the virtual power plant operating resources; The intraday market clearing result generation module is used to perform feature-guided modulation on the set of enhanced representations of the time series states of the virtual power plant's operating resources based on the virtual power plant market-related information to obtain a set of virtual power plant operating resource state representations guided by the market state, and obtain the intraday market clearing result based on the set of virtual power plant operating resource state representations guided by the market state.

10. The virtual power plant clearing management system based on multi-energy synergy according to claim 9 is characterized in that: The virtual power plant operation resource time series data encoding module includes: A virtual power plant operating resource time series data sorting unit, used to sort the load power consumption time series data of the distributed resources of the multiple virtual power plants according to the time dimension to obtain a data set of the virtual power plant distributed resource load power consumption time series input vector; A virtual power plant operating resource timing feature generation unit is used to respectively pass each virtual power plant distributed resource load power consumption timing input vector in the data set of the virtual power plant distributed resource load power consumption timing input vector through a load power consumption timing encoder based on a forward LSTM model to obtain a set of virtual power plant operating resource timing state representation vectors as the set of virtual power plant operating resource timing state representation vectors.

Images