Power data multi-dimensional aggregation method, control device and storage medium

Abstract

The embodiment of the invention provides an electric power data multi-dimensional aggregation method, a control device and a storage medium, and belongs to the technical field of electric power. The method comprises the steps of obtaining a green index value of a target power system based on data related to power generation; obtaining a profit index value A of the target power system based on the data related to the profit; through normalization processing, the green index value is converted into a corresponding profit index value B; distributing the profit index value A and the profit index value B to each data sample set in the total data sample set by using a cooperative game method; and aggregating each data sample set by using a preset aggregation algorithm to obtain a value aggregation model corresponding to each data sample set so as to determine an aggregation value of each data sample set. According to the embodiment of the invention, the contribution value of the single sample to the data set is added, and the index value is reasonably distributed by using the shapley method, so that power generation resources can be more effectively distributed, and the reliability and practicability of power supply are ensured.

Description

Technical Field

[0001] The present invention relates to the field of electric power technology, and in particular to a method for multi-dimensional aggregation of electric power data, a control device and a storage medium. Background Art

[0002] The power market is a complex multi-agent system, which involves multiple participants such as power generation companies, grid operators, energy service providers, and consumers. With the high proportion of renewable energy access and the integration of multi-source heterogeneous spatiotemporal data outside the power market, the application scenarios of the power market have risen from traditional typical scenarios to massive operation scenarios, and the traditional single-objective decision-making process has been transformed into a multi-objective decision-making process that adapts to safety, greenness, and economy. This has greatly increased the difficulty of quantifying data value. Since the multi-dimensional data related to electricity combines information from different sources, departments, and levels to provide a more comprehensive perspective, it helps to avoid one-sided perspectives and decisions based on incomplete information, and by analyzing data from multiple dimensions, hidden patterns, trends, and associations can be discovered. Therefore, it is necessary to shift from the quantification of single-dimensional data value to the quantification of multi-dimensional data, and to establish relevant value aggregation models to achieve rapid processing and analysis of massive power market data, dig out more value, and promote the digital transformation of the power industry.

[0003] At present, the commonly used value aggregation methods include weighted average method, maximum method, minimum method, median method, similarity-based aggregation method, hierarchical aggregation method, etc. The shortcomings of existing value aggregation methods include: 1) Most of the above methods are completed using simple mathematical calculations, which are relatively superficial and cannot be used to deeply calculate the aggregate value of power market data; 2) The correlation between each sample is not considered. The above method simply calculates a single value without considering its contribution to the sample set. Summary of the invention

[0004] An object of an embodiment of the present invention is to provide a method for multi-dimensional aggregation of power data, which can deeply calculate the aggregation value of power market data.

[0005] In order to achieve the above-mentioned purpose, an embodiment of the present invention provides a method for multi-dimensional aggregation of power data, which comprises: taking the thermal power units of the target power system as units, obtaining data related to power generation of each thermal power unit, constructing an actual sample set of each thermal power unit according to the data related to power generation, and obtaining a predicted sample set of each thermal power unit by using a preset power load prediction model; training a preset power dispatching model by using the actual sample sets and predicted sample sets of multiple thermal power units, wherein a trained power dispatching model is obtained by setting the objective function and constraint conditions of the preset power dispatching model, and is used to output the green index value of the target power system; obtaining data related to profit, and obtaining a predicted sample set of each thermal power unit according to the predicted sample set of the preset power load prediction model; and data, establish a profit sample set, and use the profit sample set to train a preset electricity sales profit model to obtain a trained electricity sales profit model, which is used to output a profit index value A of the target power system; convert the green index value into a corresponding profit index value B through normalization processing; use a cooperative game method to assign the profit index value A and the profit index value B to each data sample set in the total data sample set, wherein the power data is obtained through a third-party system to construct the total data sample set; and after reallocating weights to the green index and the profit index, use a preset aggregation algorithm to aggregate each data sample set to obtain a value aggregation model corresponding to each data sample set to determine the aggregation value of each data sample set.

[0006] Optionally, the data related to power generation includes power load and renewable energy output, and the data related to profit includes real-time electricity price and date cycle characteristics.

[0007] Optionally, based on carbon emissions, the preset power dispatch model is constructed, and the objective function of the preset power dispatch model is set by the following formula:

[0008]

[0009]

[0010] Among them, E and E′ represent the total CO2 emissions of the system under the actual sample set and predicted sample set of multiple thermal power units, respectively, M represents the total number of selected samples, T represents the total scheduling time length in 1 day, and They represent the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in time period t, respectively. N represents the number of thermal power units in the system. and represents the output of thermal power unit i in the mth actual sample set and the predicted sample set in period t, a, b, c represent the coefficients of the coal consumption characteristic function of the thermal power unit, k coal Indicates the CO2 emissions generated by the combustion of unit standard coal.

[0011] Optionally, the constraint condition of the preset power dispatching model includes a load balance constraint, and the load balance constraint of the preset power dispatching model is set by the following formula:

[0012]

[0013]

[0014] Among them, D m and D′ m They represent the actual load output and predicted load output corresponding to the mth actual sample set and predicted sample set, respectively. and They represent the actual output of two renewable energy sources in the mth actual sample set and the predicted sample set in period t, and They represent the predictions of two renewable energy sources of the mth actual sample set and the predicted sample set in period t, respectively, and N represents the number of thermal power units in the system.

[0015] Optionally, the constraints of the preset power dispatch model also include upper and lower output limits of thermal power units and / or thermal power unit ramp rate constraints. The upper and lower output limits of thermal power units of the preset power dispatch model are set by the following formula:

[0016]

[0017]

[0018] in, and represents the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in period t, respectively. i,min and P i,maX They represent the minimum and maximum technical output of thermal power unit i, and They represent the actual output and predicted output of any renewable energy source in the mth actual sample set and predicted sample set in time period t, respectively.

[0019] The ramp rate constraint of the thermal power unit of the preset power dispatching model is set by the following formula:

[0020]

[0021]

[0022] Among them, α iup represents the upward climbing rate of thermal power unit i, α idown It represents the downward climbing rate of thermal power unit i, and Δt represents the time interval between period t and period t-1.

[0023] Optionally, the constraint condition of the preset power dispatching model further includes a minimum start and stop time constraint. The minimum start and stop time constraint of the preset power dispatching model is set by the following formula:

[0024]

[0025]

[0026] in, and They represent the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in period t, and They represent the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in the t-1 period, and They represent the actual continuous operation time and the predicted continuous operation time of thermal power unit i at time t-1, T ion Indicates the minimum continuous operation time of thermal power unit i, and They represent the actual continuous shutdown time and predicted continuous shutdown time of thermal power unit i at time t-1, T ioff Indicates the minimum continuous shutdown time of thermal power unit i.

[0027] Optionally, normalization processing is performed through the following formula to convert the green index value into a corresponding profit index value B:

[0028]

[0029] in, represents the profit index value B, |E-E′| represents the green index value, LCOE CCUS ×|E-E′| represents the cost of reducing the same amount of CO2 through CCUS technology, p trade Represents the carbon trading price.

[0030] Optionally, the method of using a cooperative game method to distribute the profit index value A and the profit index value B to each data sample set in the total data sample set includes: based on the cooperative game method, introducing a Shapley algorithm to distribute the profit index value A and the profit index value B to each data sample set in the total data sample set. The index value of the data sample set m under the green index or the profit index is calculated by the following formula:

[0031]

[0032] Where n represents the number of data sample sets, x represents green value or economic value, φ x (m) represents the value of data sample set m from the overall benefit v(N) under x, S m represents the set of all sample subsets containing m in a certain value calculation scenario, P x (s) represents the probability of occurrence of the corresponding cooperation sequence combination s under x, vx(s) is the total value of the data under x, and v x (s)-v x (s\{m}) represents the marginal value created by the data sample set m for participating in different sets S under x for its own participation in the alliance.

[0033] An embodiment of the present invention also provides a control device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned multi-dimensional aggregation method of power data.

[0034] An embodiment of the present invention further provides a machine-readable storage medium, on which instructions are stored, and the instructions enable a machine to execute the above-mentioned multi-dimensional aggregation method of power data.

[0035] Through the above technical scheme, the embodiment of the present invention constructs actual sample sets and predicted sample sets of multiple thermal power units based on data related to power generation, and obtains a trained power dispatch model by setting the objective function and constraint conditions of the preset power dispatch model, which is used to output the green index value of the target power system; based on the data related to profit, the preset power sales profit model is trained to obtain the trained power sales profit model, which is used to output the profit index value A of the target power system; through normalization, the green index value is converted into the corresponding profit index value B; using the cooperative game method, the profit index value A and the profit index value B are allocated to each data sample set in the total data sample set; after redistributing the weights of the green index and the profit index, each data sample set is aggregated using a preset aggregation algorithm to obtain the value aggregation model corresponding to each data sample set to determine the aggregation value of each data sample set. The aggregation method provided by the embodiment of the present invention can integrate scattered, isolated, and heterogeneous data resources to form a data resource with a higher aggregation value, thereby providing better data support and services for the power system. The embodiment of the present invention adds the contribution value of a single sample to the data set, and uses the Shapley method to reasonably allocate the index values, and divides the electricity market into two categories: green index values ​​and profit index values. Through the aggregation method provided by the embodiment of the present invention, power generation resources can be allocated more effectively to ensure the reliability and practicality of power supply.

[0036] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0038] Figure 1 is a flow chart of a method for multi-dimensional aggregation of power data provided by an embodiment of the present invention; and

[0039] Figure 2 It is a flowchart of an example multi-dimensional aggregation method of power data. DETAILED DESCRIPTION

[0040] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0041] Before explaining the embodiments of the present invention in detail, the terms involved in the embodiments of the present invention are first explained.

[0042] Value aggregation: Aggregation of multi-source spatiotemporal data means combining data from different sources and different time points to work together in the decision-making model. This aggregation is not just a simple superposition of data, but it must fully consider the correlation between data, time series characteristics, and their practical significance in specific scenarios.

[0043] Data aggregation in the power market: refers to the integration of different power resources, consumer demand, market price signals and other relevant factors in the power market to optimize the allocation of power resources and the operation of the power system.

[0044] Cooperative game: a special form of game in which participants play the game through cooperation or alliance, aiming to maximize collective interests rather than just pursuing individual interests. This form of game emphasizes the importance of cooperation and believes that a win-win situation can be achieved through cooperation, that is, the interests of all participants can be increased, or at least the interests of one party can be increased while the interests of the other party are not harmed, so that the interests of the entire society can also be increased.

[0045] Shapley value method: a method to measure the contribution of each participant to the collective benefit in a cooperative game. This method takes into account the order in which each participant joins the cooperation and calculates the contribution value of each participant to the final result.

[0046] Optimization scheduling model: An applied mathematical model used to solve resource allocation and scheduling problems to achieve the optimization of specific goals. This model is widely used in economic management work, including but not limited to linear programming, nonlinear programming, dynamic programming, integer programming and system science methods. It is an important tool that helps decision makers find the optimal resource allocation and scheduling plan under specific conditions through mathematical methods and algorithms to maximize economic and social benefits.

[0047] Figure 1 This is a flow chart of the multi-dimensional aggregation method of power data provided by an embodiment of the present invention. Please refer to Figure 1 , the multi-dimensional aggregation method of power data may include the following steps:

[0048] Step S110: Taking the thermal power units of the target power system as units, obtaining data related to power generation of each thermal power unit, constructing an actual sample set of each thermal power unit based on the data related to power generation, and obtaining a predicted sample set of each thermal power unit using a preset power load prediction model.

[0049] In the embodiment of the present invention, the data related to power generation may preferably include power load and renewable energy output, etc.

[0050] Please combine Figure 2 For example, taking the thermal power units of the target power system as units, the data related to the power generation of each thermal power unit is obtained, and the actual sample set of each thermal power unit is constructed based on the data related to the power generation. For example, the hourly power load, renewable energy output (for example, wind power output, photovoltaic output, etc.) and total capacity of thermal power installed units in a certain place from April to October 2023 provided by a power trading center are selected, and the actual sample set of each thermal power unit is constructed based on the obtained data, and the actual sample sets of multiple thermal power units are formed, that is, the total actual sample set.

[0051] Since the embodiment of the present invention preferably integrates the prediction data in the power dispatching model, the prediction sample set of each thermal power unit can be obtained by using the preset power load prediction model. The preset power load prediction model can use the RBF model to predict the power load.

[0052] Step S120: Use actual sample sets and predicted sample sets of multiple thermal power units to train a preset power dispatching model, wherein, by setting the objective function and constraints of the preset power dispatching model, a trained power load prediction model and a trained power dispatching model are obtained, which are used to output the green index value of the target power system.

[0053] To illustrate with an example, for the green index value (also called green value), a multi-dimensional value aggregation model based on electricity market data (i.e., a power dispatch model constructed based on carbon emissions) is established, and its decision variables are, for example, the power of the thermal power unit at each moment, and the objective function is to minimize the carbon emissions of the thermal power unit. In order to describe the index (or value) of the actual sample set, it is necessary to obtain the initial index when the actual sample set is not used, that is, the baseline value of carbon emissions (for example, a typical day is selected), and then the carbon emissions calculated using the actual sample set are subtracted from the baseline value to obtain the index value of the actual sample set (i.e., the green index value). Therefore, the embodiment of the present invention uses a preset power load forecasting model to obtain a forecast sample set for each thermal power unit.

[0054] Preferably, based on carbon emissions, the preset power dispatch model is constructed, and the objective function of the preset power dispatch model is set by the following formula:

[0055]

[0056]

[0057] Among them, E and E′ represent the total CO2 emissions of the system under the actual sample set and predicted sample set of multiple thermal power units, respectively, M represents the total number of selected samples (samples can be divided by days), T represents the total scheduling time length in 1 day (for example, the time step is 1 hour), and They represent the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in time period t, respectively. N represents the number of thermal power units in the system. and represents the output of thermal power unit i in the mth actual sample set and the predicted sample set in the tth period, a, b, c represent the coefficients of the coal consumption characteristic function of the thermal power unit (which can be obtained by measuring and fitting the operating data of the thermal power unit), k coal It indicates the CO2 emissions produced by burning a unit of standard coal. For example, burning a unit of standard coal produces 2.7 tons of carbon dioxide.

[0058] In the embodiment of the present invention, the constraint condition of the preset power dispatching model preferably includes a load balance constraint. The load balance constraint of the preset power dispatching model can be set by the following formula:

[0059]

[0060]

[0061] Among them, D m and D′ m They represent the actual load output and predicted load output corresponding to the mth actual sample set and predicted sample set, respectively. and They represent the actual output of two renewable energy sources (e.g., wind power and photovoltaic power) in the mth actual sample set and the predicted sample set in period t, respectively. and They represent the predictions of two renewable energy sources of the mth actual sample set and the predicted sample set in period t, respectively, and N represents the number of thermal power units in the system.

[0062] The load balance constraint is one of the most basic constraints for the safe and stable operation of the power system. The load balance constraint of the embodiment of the present invention requires that under M samples, the sum of the total power generated by the participating systems at each moment is equal to the total load of the system under M samples.

[0063] The constraints of the preset power dispatch model in the preferred embodiment of the present invention may also include upper and lower output limits of thermal power units and / or thermal power unit ramp rate constraints. Preferably, the upper and lower output limits of the thermal power units of the preset power dispatch model may be set by the following formula:

[0064]

[0065]

[0066] in, and represents the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in the tth period, P i,min and P i,max They represent the minimum and maximum technical output of thermal power unit i, and They represent the actual output and predicted output of any renewable energy source in the mth actual sample set and predicted sample set in time period t, respectively.

[0067] Preferably, the ramp rate constraint of the thermal power unit of the preset power dispatching model can be set by the following formula:

[0068]

[0069]

[0070] Among them, α iup represents the upward climbing rate of thermal power unit i (% / min), α idown represents the downward ramp rate of thermal power unit i (% / min), and Δt represents the time interval between period t and period t-1. The thermal power unit ramp rate constraint defines the upper and lower limits of the thermal power unit's regulation capacity per unit time.

[0071] The constraint condition of the preset power dispatching model in the embodiment of the present invention may also include a minimum start and stop time constraint. Preferably, the minimum start and stop time constraint of the preset power dispatching model may be set by the following formula:

[0072]

[0073]

[0074] in, and They represent the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in period t, and They represent the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in the t-1 period, and They represent the actual continuous operation time and the predicted continuous operation time of thermal power unit i at time t-1, T ion Indicates the minimum continuous operation time of thermal power unit i, and They represent the actual continuous shutdown time and predicted continuous shutdown time of thermal power unit i at time t-1, T ioff It represents the minimum continuous downtime of thermal power unit i. The minimum start-stop time constraint is an operation constraint caused by the equipment characteristics of the thermal power unit.

[0075] Following the above example, based on the objective function and constraint conditions of the preset power dispatch model set above, a trained power dispatch model is obtained. Based on the above power dispatch model, the carbon emissions of M actual sample sets and predicted sample sets can be obtained through the solver, so that the green index value (i.e., the total green value) of the M samples is obtained as |E-E′|.

[0076] Step S130: Acquire profit-related data, establish a profit sample set based on the profit-related data, and use the profit sample set to train a preset electricity sales profit model to obtain a trained electricity sales profit model for outputting a profit index value A of the target power system.

[0077] The profit-related data in the embodiment of the present invention may preferably include real-time electricity prices and date cycle characteristics, etc.

[0078] Please refer to Figure 2 As in the example of step S110, the real-time electricity price and date cycle characteristics of a certain place from April to October 2023 provided by a certain power trading center are selected to establish a profit sample set. For the profit index value, for example, a power sales profit model based on the day-ahead electricity price forecast is established to measure the profit index value A of the target power system. Among them, the power sales profit model is as follows:

[0079] For example, predict the real-time electricity price y = [y0,y1,…,y 24 ] T , it is known that the hourly renewable energy output (e.g., hydropower, wind power, photovoltaic power, etc.) predicted by meteorological data for the day is and the periodic characteristics of today's date (e.g., month, day of the week)X p =[x m ,x w ] T ,make Then, we can use the known data to build a machine learning model to predict electricity prices, as shown below:

[0080]

[0081] For power sales companies, the electricity price predicted by the above model is The real electricity price y i The closer the distance is, the higher the profit the power sales company can obtain, and vice versa. Consider the economic profit of the power sales company Among them, w i is the maximum profit that the electricity sales company can obtain when the hourly forecast is accurate. They are all normal values. SMAPE is the symmetric mean absolute percentage error, which is defined as follows:

[0082]

[0083] When there is no data, the electricity sales company can use the average electricity price for each hour of the year. As the predicted value of daily electricity price, the revenue is After using the trained electricity sales profit model, the profit obtained by using the predicted electricity price is Therefore, for the dataset D S ={X i} i∈S , let f S To utilize D S The profit index value of the trained model is

[0084] Step S140: convert the green index value into a corresponding profit index value B through normalization processing.

[0085] As an example, the indicator values ​​of the above different dimensions (green and profit) are dimensionalized and the values ​​of each dimension are unified through the profit dimension. For the green indicator value, the benefits brought by the reduction of carbon emissions can be used to describe it. The green indicator value of the data can be regarded as the cost of reducing the same amount of CO2 by means of bioenergy and carbon capture and storage (CCUS) technology.

[0086] Preferably, the green index value may be converted into a corresponding profit index value B by normalization processing using the following formula:

[0087]

[0088] in, represents the profit index value B, |E-E′| represents the green index value, LCOE CCUS ×|E-E′| represents the cost of reducing the same amount of CO2 through CCUS technology, p trade Represents the carbon trading price.

[0089] Step S150: using a cooperative game method, the profit index value A and the profit index value B are allocated to each data sample set in the total data sample set, wherein the power data is acquired through a third-party system to construct the total data sample set.

[0090] For example, the third-party system is an electricity trading center. The electricity data of a certain place from April to September 2023 is obtained through the third-party system to construct a total set of data samples.

[0091] Preferably, the method of using the cooperative game method to distribute the profit index value A and the profit index value B to each data sample set in the total data sample set may include: based on the cooperative game method, introducing the Shapley algorithm to distribute the profit index value A and the profit index value B to each data sample set in the total data sample set, and calculating the index value of the data sample set m under the green index or the profit index by the following formula:

[0092]

[0093] Where n represents the number of data sample sets, x represents green value or economic value, φ x (m) represents the value of data sample set m from the overall benefit v(N) under x, S m represents the set of all sample subsets containing m in a certain value calculation scenario, P x (s) represents the probability of occurrence of the corresponding cooperation sequence combination s under x, vx(s) is the total value of the data under x, and v x (s)-v x (s\{m}) represents the marginal value created by the data sample set m for participating in different sets S under x for its own participation in the alliance.

[0094] Step S160: After reallocating the weights of the green index and the profit index, each data sample set is aggregated using a preset aggregation algorithm to obtain a value aggregation model corresponding to each data sample set to determine the aggregation value of each data sample set.

[0095] Among them, the weights of different dimensions (green indicators and profit indicators) can be determined by the entropy weight method.

[0096] Take an example to illustrate that weights are assigned to indicators of different dimensions, and the values ​​of each dimension of each sample are aggregated for analysis. Based on the additivity of the Shapley algorithm, the value aggregation model corresponding to each data sample set can be obtained. The aggregate value of each data sample set can be calculated using the following model:

[0097]

[0098] Where φ(i) is the weighted total value of sample m under two scenarios (green value or economic value, for example, x = 1 represents the scenario under the green indicator, and x = 2 represents the scenario under the profit indicator), qm (i) is the weight value of sample m in scene i, φ m (i) is the aggregate value of sample m in scene i.

[0099] Continuing with the above example, based on formula (15), after calculating the aggregation value of each data sample set, the data can be divided into high aggregation value and low aggregation value, so as to evaluate the index value of the power industry data.

[0100] Accordingly, the embodiment of the present invention constructs actual sample sets and predicted sample sets of multiple thermal power units based on data related to power generation, and obtains a trained power dispatch model by setting the objective function and constraint conditions of the preset power dispatch model, which is used to output the green index value of the target power system; based on the data related to profit, the preset power sales profit model is trained to obtain the trained power sales profit model, which is used to output the profit index value A of the target power system; through normalization, the green index value is converted into the corresponding profit index value B; using the cooperative game method, the profit index value A and the profit index value B are allocated to each data sample set in the total data sample set; after redistributing the weights of the green index and the profit index, each data sample set is aggregated using the preset aggregation algorithm to obtain the value aggregation model corresponding to each data sample set to determine the aggregation value of each data sample set. The aggregation method provided by the embodiment of the present invention can integrate scattered, isolated, and heterogeneous data resources together to form a data resource with more aggregated value, thereby providing better data support and services for the power system. The embodiment of the present invention adds the contribution value of a single sample to the data set, and uses the Shapley method to reasonably allocate the index values, and divides the electricity market into two categories: green index values ​​and profit index values. Through the aggregation method provided by the embodiment of the present invention, power generation resources can be allocated more effectively to ensure the reliability and practicality of power supply.

[0101] An embodiment of the present invention also provides a control device, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned multi-dimensional aggregation method of power data.

[0102] An embodiment of the present invention further provides a machine-readable storage medium, on which instructions are stored, and the instructions enable a machine to execute the above-mentioned multi-dimensional aggregation method of power data.

[0103] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0104] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0105] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0106] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0107] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0108] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0109] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0110] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0111] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A multi-dimensional aggregation method for power data, characterized in that: The multi-dimensional aggregation method of power data includes: Taking the thermal power units of the target power system as units, obtaining data related to power generation of each thermal power unit, constructing an actual sample set of each thermal power unit according to the data related to power generation, and obtaining a predicted sample set of each thermal power unit by using a preset power load prediction model; Using actual sample sets and predicted sample sets of multiple thermal power units, a preset power dispatching model is trained, wherein, by setting the objective function and constraint conditions of the preset power dispatching model, a trained power dispatching model is obtained, which is used to output the green index value of the target power system; Acquire profit-related data, establish a profit sample set based on the profit-related data, and use the profit sample set to train a preset electricity sales profit model to obtain a trained electricity sales profit model for outputting a profit index value A of the target power system; By normalization processing, the green index value is converted into a corresponding profit index value B; Using a cooperative game method, the profit index value A and the profit index value B are allocated to each data sample set in the total data sample set, wherein the power data is acquired through a third-party system to construct the total data sample set; and After reallocating the weights of the green index and the profit index, each data sample set is aggregated using a preset aggregation algorithm to obtain a value aggregation model corresponding to each data sample set to determine the aggregation value of each data sample set.

2. The method for multi-dimensional aggregation of power data according to claim 1, characterized in that: The data related to power generation includes power load and renewable energy output, and the data related to profit includes real-time electricity price and date cycle characteristics.

3. The method for multi-dimensional aggregation of power data according to claim 1, characterized in that: Based on carbon emissions, the preset power dispatch model is constructed, and the objective function of the preset power dispatch model is set by the following formula: Among them, E and E′ represent the total CO2 emissions of the system under the actual sample set and predicted sample set of multiple thermal power units, respectively, M represents the total number of selected samples, T represents the total scheduling time length in 1 day, and They represent the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in time period t, respectively. N represents the number of thermal power units in the system. and represents the output of thermal power unit i in the mth actual sample set and the predicted sample set in period t, a, b, c represent the coefficients of the coal consumption characteristic function of the thermal power unit, k coal Indicates the CO2 emissions generated by the combustion of unit standard coal.

4. The method for multi-dimensional aggregation of power data according to claim 1, characterized in that: The constraint condition of the preset power dispatching model includes a load balance constraint. The load balance constraint of the preset power dispatching model is set by the following formula: Among them, D m and D′ m They represent the actual load output and predicted load output corresponding to the mth actual sample set and predicted sample set, respectively. and They represent the actual output of two renewable energy sources in the mth actual sample set and the predicted sample set in period t, and They represent the predictions of two renewable energy sources of the mth actual sample set and the predicted sample set in period t, respectively, and N represents the number of thermal power units in the system.

5. The method for multi-dimensional aggregation of power data according to claim 4, characterized in that: The constraints of the preset power dispatch model also include upper and lower limits of thermal power unit output and / or thermal power unit ramp rate constraints. The upper and lower limits of the thermal power unit output of the preset power dispatch model are set by the following formula: in, and represents the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in the tth period, P i，min and P i，max They represent the minimum and maximum technical output of thermal power unit i, and They represent the actual output and predicted output of any renewable energy source in the mth actual sample set and predicted sample set in period t, respectively. The ramp rate constraint of the thermal power unit of the preset power dispatching model is set by the following formula: Among them, α iup represents the upward climbing rate of thermal power unit i, α idown It represents the downward climbing rate of thermal power unit i, and Δt represents the time interval between period t and period t-1.

6. The method for multi-dimensional aggregation of power data according to claim 4, characterized in that: The constraint condition of the preset power dispatching model also includes a minimum start and stop time constraint. The minimum start and stop time constraint of the preset power dispatching model is set by the following formula: in, and They represent the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in period t, and They represent the operating status of thermal power unit i in the mth actual sample set and the predicted sample set in the t-1 period, and They represent the actual continuous operation time and the predicted continuous operation time of thermal power unit i at time t-1, T ion Indicates the minimum continuous operation time of thermal power unit i, and They represent the actual continuous shutdown time and predicted continuous shutdown time of thermal power unit i at time t-1, T ioff Indicates the minimum continuous shutdown time of thermal power unit i.

7. The method for multi-dimensional aggregation of power data according to claim 1, characterized in that: The green index value is converted into the corresponding profit index value B by normalization processing using the following formula: in, represents the profit index value B, |EE′| represents the green index value, LCOE CCUS ×|EE′| represents the cost of reducing the same amount of CO2 through CCUS technology, p trade Represents the carbon trading price.

8. The method for multi-dimensional aggregation of power data according to claim 1, characterized in that: The method of using the cooperative game method to distribute the profit index value A and the profit index value B to each data sample set in the total data sample set includes: Based on the cooperative game method, the Shapley algorithm is introduced to distribute the profit index value A and the profit index value B to each data sample set in the total data sample set. The index value of the data sample set m under the green index or profit index is calculated by the following formula: Where n represents the number of data sample sets, x represents green value or economic value, φ x (m) represents the value of data sample set m from the overall benefit v(N) under x, S m represents the set of all sample subsets containing m in a certain value calculation scenario, P x (s) represents the probability of occurrence of the corresponding cooperation sequence combination s under x, v x (s) is the total value of data under x, v x (s)-v x (s\{m}) represents the marginal value created by the data sample set m for participating in different sets S under x for its own participation in the alliance.

9. A control device, characterized in that: The control device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the multi-dimensional aggregation method for power data according to any one of claims 1-8.

10. A machine-readable storage medium, characterized in that: The machine-readable storage medium stores instructions, which enable the machine to execute the multi-dimensional aggregation method for power data according to any one of claims 1-8.

Images