Virtual power plant active real-time scheduling equivalent model construction method and active scheduling method

Abstract

The application discloses a virtual power plant active real-time scheduling equivalent model construction method and an active scheduling method. The virtual power plant active real-time scheduling equivalent model construction method comprises the following steps: obtaining a plurality of training active real-time scheduling instructions; obtaining an internal active real-time optimization scheduling model of a virtual power plant; determining a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results corresponding to the internal active real-time optimization scheduling model of the virtual power plant and the plurality of training active real-time scheduling instructions; training a first neural network based on the plurality of training active real-time scheduling instructions and the plurality of training real-time scheduling costs, to obtain an active real-time scheduling cost model; and training a second neural network based on the plurality of training active real-time scheduling instructions and the plurality of training scheduling instruction feasibility analysis results, to obtain an active real-time scheduling instruction feasibility model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power grid, and particularly relates to a virtual power plant active real-time scheduling equivalent model construction method and an active scheduling method. BACKGROUND

[0002] For a given grid active real-time scheduling instruction, in order to minimize the grid active real-time scheduling cost, the virtual power plant with a small real-time scheduling cost should be scheduled preferentially. Therefore, in the case that the day-ahead generation plan of each virtual power plant has been made, for a given grid active real-time scheduling instruction, determining the active real-time scheduling cost of each virtual power plant is a problem to be solved urgently. SUMMARY

[0003] Therefore, the embodiments of the present application provide a virtual power plant active real-time scheduling equivalent model construction method and an active scheduling method.

[0004] According to a first aspect of the present application, the embodiments of the present application provide a virtual power plant active real-time scheduling equivalent model construction method, comprising:

[0005] A plurality of training active real-time scheduling instructions are obtained, wherein the training active real-time scheduling instruction comprises a training scheduling start period, a training scheduling duration and a training scheduling power;

[0006] A virtual power plant internal active real-time optimization scheduling model is obtained; the virtual power plant internal active real-time optimization scheduling model meets the requirements of the training active real-time scheduling instruction, and determines the planned output power curve of each type of distributed power source in the virtual power plant within a scheduling period, and takes the minimum real-time scheduling cost of the virtual power plant within a scheduling period as the target;

[0007] Based on the virtual power plant internal active real-time optimization scheduling model and the plurality of training active real-time scheduling instructions, a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results corresponding to the plurality of training real-time scheduling costs are determined;

[0008] Based on the plurality of training active real-time scheduling instructions and the plurality of training real-time scheduling costs, a first neural network is trained to obtain an active real-time scheduling cost model;

[0009] Based on the plurality of training active real-time scheduling instructions and the plurality of training scheduling instruction feasibility analysis results, a second neural network is trained to obtain an active real-time scheduling instruction feasibility model; based on the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model, the real-time scheduling cost and the scheduling instruction feasibility analysis result of the virtual power plant executing the grid active real-time scheduling instruction are determined.

[0010] Optionally, obtaining the virtual power plant internal active real-time optimization scheduling model comprises:

[0011] obtaining a target function corresponding to the active real-time optimization scheduling model inside the virtual power plant; the target function is related to a first operation cost function of the heat power source inside the virtual power plant in a scheduling period, and a second operation cost function of the energy storage system power source in the scheduling period;

[0012] determining a constraint condition corresponding to the target function; the constraint condition includes a first operation constraint condition corresponding to the heat power source, a second operation constraint condition corresponding to the energy storage system power source, and a power balance constraint condition inside the virtual power plant.

[0013] Optionally, the power balance constraint condition inside the virtual power plant includes:

[0014] In the training scheduling duration, the sum of the output power of the heat power source inside the virtual power plant, the output power of the energy storage system power source, and the training scheduling power corresponding to the training active real-time scheduling instruction, is equal to the total power of the load inside the virtual power plant.

[0015] Optionally, the plurality of training active real-time scheduling instructions are obtained, including:

[0016] determining a scheduling start time range, a scheduling duration range, and a scheduling power range corresponding to the training active real-time scheduling instruction;

[0017] based on the scheduling start time range, the scheduling duration range, and the scheduling power range, a plurality of training active real-time scheduling instructions are randomly generated by using a uniform distribution function.

[0018] Optionally, based on the active real-time optimization scheduling model inside the virtual power plant and the plurality of training active real-time scheduling instructions, a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results are determined, including:

[0019] based on each training active real-time scheduling instruction, the active real-time optimization scheduling model inside the virtual power plant is solved to obtain a solution result; the solution result includes a first output power curve of the heat power source inside the virtual power plant in the training scheduling duration, a second output power curve of the energy storage system power source in the training scheduling duration, and a training minimum real-time scheduling cost in a corresponding scheduling period;

[0020] if the solution result shows that the first output power curve, the second output power curve, and the training minimum real-time scheduling cost have a solution, the training minimum real-time scheduling cost is taken as the corresponding training real-time scheduling cost; and the training active scheduling instruction feasibility analysis result is determined to be feasible;

[0021] if the solution result shows that the first output power curve, the second output power curve, or the training minimum real-time scheduling cost has no solution, the training active scheduling instruction feasibility analysis result is determined to be infeasible.

[0022] According to the second aspect of the application, the embodiments of the application provide a virtual power plant active power scheduling method, comprising:

[0023] obtaining an active power scheduling instruction of a power grid to the virtual power plant, the active power scheduling instruction comprising a scheduling start period, a scheduling duration period and a scheduling power;

[0024] processing the active power scheduling instruction based on an active power real-time scheduling cost model and an active power real-time scheduling instruction feasibility model corresponding to each virtual power plant, to obtain a real-time scheduling cost and a scheduling instruction feasibility analysis result of each virtual power plant when executing the active power scheduling instruction; the active power real-time scheduling cost model and the active power real-time scheduling instruction feasibility model are constructed by using the virtual power plant active power real-time scheduling equivalent model construction method in the first aspect or any of the embodiments of the first aspect;

[0025] determining a target virtual power plant based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, and using the target virtual power plant to execute the active power scheduling instruction.

[0026] Optionally, determining the target virtual power plant based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant comprises:

[0027] determining a virtual power plant with the smallest real-time scheduling cost and the most feasible scheduling instruction feasibility analysis result as the target virtual power plant based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant.

[0028] According to the third aspect of the application, the embodiments of the application provide a virtual power plant active power real-time scheduling equivalent model construction device, comprising:

[0029] a first obtaining module configured to obtain a plurality of training active power real-time scheduling instructions, the training active power real-time scheduling instruction comprising a training scheduling start period, a training scheduling duration period and a training scheduling power;

[0030] a second obtaining module configured to obtain a virtual power plant internal active power real-time optimization scheduling model; the virtual power plant internal active power real-time optimization scheduling model meets the requirements of the training active power real-time scheduling instruction, and determines a planned output power curve of each type of distributed power source in the virtual power plant within a scheduling period, and takes the minimum real-time scheduling cost of the virtual power plant within a scheduling period as the target;

[0031] a first determining module configured to determine a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results based on the virtual power plant internal active power real-time optimization scheduling model and the plurality of training active power real-time scheduling instructions;

[0032] The first training module is configured to train the first neural network based on the plurality of training active real-time scheduling instructions and the corresponding plurality of training real-time scheduling costs, and obtain an active real-time scheduling cost model.

[0033] The second training module is configured to train the second neural network based on the plurality of training active real-time scheduling instructions and the corresponding plurality of training scheduling instruction feasibility analysis results, and obtain an active real-time scheduling instruction feasibility model.

[0034] According to a fourth aspect of the present application, an embodiment of the present application provides a virtual power plant active scheduling device, comprising:

[0035] The third obtaining module is configured to obtain an active scheduling instruction of a power grid to the virtual power plant, the active scheduling instruction comprising a scheduling start time period, a scheduling duration time period and a scheduling power;

[0036] The processing module is configured to process the active scheduling instruction based on the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model corresponding to each virtual power plant, and obtain the real-time scheduling cost and the scheduling instruction feasibility analysis result of the active scheduling instruction executed by each virtual power plant; the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model are constructed by using the virtual power plant active real-time scheduling equivalent model construction method in the first aspect or any implementation manner of the first aspect;

[0037] The second determining module is configured to determine a target virtual power plant based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, and execute the active scheduling instruction by using the target virtual power plant.

[0038] According to a fifth aspect of the present application, an embodiment of the present application provides an electronic device, comprising:

[0039] The at least one processor and the memory connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the virtual power plant active real-time scheduling equivalent model construction method in the first aspect or any implementation manner of the first aspect, or the virtual power plant active scheduling method in the second aspect or any implementation manner of the second aspect.

[0040] The virtual power plant active real-time scheduling equivalence model construction method and the active real-time scheduling method provided by the embodiments of the present application comprise the following steps: obtaining a plurality of training active real-time scheduling instructions; obtaining an internal active real-time optimization scheduling model of the virtual power plant; determining a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results based on the internal active real-time optimization scheduling model of the virtual power plant and the plurality of training active real-time scheduling instructions; training a first neural network based on the plurality of training active real-time scheduling instructions and the plurality of training real-time scheduling costs to obtain an active real-time scheduling cost model; training a second neural network based on the plurality of training active real-time scheduling instructions and the plurality of training scheduling instruction feasibility analysis results to obtain an active real-time scheduling instruction feasibility model; in this way, when the day-ahead generation plan of the virtual power plant has been formulated, the real-time scheduling cost and the scheduling instruction feasibility analysis result of the virtual power plant when executing the grid active real-time scheduling instruction can be determined based on the constructed active real-time scheduling cost model and the active real-time scheduling instruction feasibility model, so that the active real-time scheduling cost of each virtual power plant can be quickly determined for a given grid active real-time scheduling instruction, and the grid can facilitate the active real-time scheduling of the virtual power plant.

[0041] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the embodiments of the present application can be implemented in accordance with the content of the description, and in order to make the above and other purposes, characteristics and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 The flowchart of the virtual power plant active real-time scheduling equivalence model construction method in the embodiments of the present application is shown in the figure;

[0043] Figure 2 The flowchart of the virtual power plant active real-time scheduling equivalence model construction method in the embodiments of the present application is shown in the figure;

[0044] Figure 3 The structure diagram of the virtual power plant active real-time scheduling equivalence model construction device in the embodiments of the present application is shown in the figure;

[0045] Figure 4 The structure diagram of the virtual power plant active real-time scheduling equivalence model construction device in the embodiments of the present application is shown in the figure;

[0046] Figure 5 The hardware structure diagram of the electronic device in the embodiments of the present application is shown in the figure. DETAILED DESCRIPTION

[0047] In order to make the purposes, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0048] Since the types of distributed power sources contained in each virtual power plant are different, each type of distributed power source has its privacy protection demand, and each virtual power plant usually does not disclose the unit composition, physical model, and operation information of each distributed power source inside it. Therefore, the grid dispatching institution can only obtain the dispatching capacity and dispatching cost information reported by each virtual power plant based on a given electricity price. Thus, the grid dispatching institution can determine the power generation plan of each virtual power plant in a dispatching period based on the dispatching capacity and dispatching cost information reported by each virtual power plant, for example, determine the planned output power curve of each type of distributed power source of each virtual power plant in a dispatching period and the power curve of power purchased from an external grid. Therefore, in the case of knowing the day-ahead power generation plan of a virtual power plant, in order to facilitate the global active real-time optimization dispatching of the grid dispatching institution, for the grid active real-time dispatching instruction, for example, [33, 8, 0.76], that is, from 8 o'clock, the virtual power plant is required to increase the power purchased from the external grid by 0.76 MW within 2 hours, each virtual power plant should provide the real-time dispatching cost corresponding to any time and any dispatching power of the next day, that is, the active real-time dispatching equivalent model of the virtual power plant.

[0049] To this end, the embodiments of the present application provide a virtual power plant active real-time dispatching equivalent model construction method, applied to a virtual power plant, and the method comprises:

[0050] S101, a plurality of training active real-time dispatching instructions are obtained, and the training active real-time dispatching instruction comprises a training dispatching start period, a training dispatching duration period, and a training dispatching power.

[0051] In the present embodiment, the virtual power plant can include wind power sources, hydroelectric power sources, photovoltaic power sources, energy storage system power sources, thermal power sources, and loads. The thermal power source can be composed of multiple groups of micro gas turbines.

[0052] In the embodiment, the virtual power plant active real-time scheduling equivalence model is: in the case that the day-ahead scheduling output power curve of each type of distributed power source of the virtual power plant and the power curve of the power purchased from the external power grid are known, the mapping model between the active real-time scheduling instruction of the next day period and the internal real-time scheduling cost of the virtual power plant. Therefore, in order to construct the virtual power plant active real-time scheduling equivalence model, a plurality of training active real-time scheduling instructions can be obtained first. The training scheduling power can be the power purchased from the external power grid or the output power to the external power grid in the training scheduling duration. When the training scheduling power is positive, it means that the power purchased from the external power grid is increased; when the training scheduling power is negative, it means that the output power to the external power grid is increased.

[0053] S102, obtain the internal active real-time optimization scheduling model of the virtual power plant; the internal active real-time optimization scheduling model of the virtual power plant takes the minimum real-time scheduling cost of the virtual power plant in a scheduling period as the target, under the conditions that the training active real-time scheduling instruction is met and the planned output power curve of each type of distributed power source in the virtual power plant in a scheduling period is determined.

[0054] In the embodiment, under a certain time section, the real-time scheduling capability of the virtual power plant is determined by the scheduling capability of each unit contained therein, and the real-time scheduling cost is determined by the marginal cost when a certain power is adjusted at the current operating point of each unit inside. In a certain time period, the scheduling behavior occurring at a certain time section can indirectly affect the scheduling capability and scheduling cost of other time sections in the future, so the real-time scheduling capability and scheduling cost of the virtual power plant should not be considered only at the current time, but should be determined by comprehensive optimization in a scheduling period. Therefore, under the condition that the day-ahead generation plan of the virtual power plant has been made, for a given grid active real-time scheduling instruction, the virtual power plant should obtain the corresponding scheduling cost by minimizing the real-time scheduling cost of the internal units in a scheduling period. Therefore, the virtual power plant can determine the minimum real-time scheduling cost in a scheduling period by constructing an internal active real-time optimization scheduling model of the virtual power plant, which takes the minimum real-time scheduling cost of the virtual power plant in a scheduling period as the target, under the conditions that the training active real-time scheduling instruction is met and the planned output power curve of each type of distributed power source in the virtual power plant in a scheduling period is determined.

[0055] S103, determine the corresponding plurality of training real-time scheduling costs and the plurality of training scheduling instruction feasibility analysis results based on the internal active real-time optimization scheduling model of the virtual power plant and the plurality of training active real-time scheduling instructions.

[0056] In the embodiment, the internal active real-time optimization scheduling model of the virtual power plant can be solved by a plurality of training active real-time scheduling instructions, to obtain the corresponding plurality of training real-time scheduling costs and the plurality of training scheduling instruction feasibility analysis results.

[0057] S104, training the first neural network based on the plurality of training active real-time scheduling instructions and the corresponding plurality of training real-time scheduling costs to obtain an active real-time scheduling cost model.

[0058] In the embodiment, the plurality of training active real-time scheduling instructions can be used as training data, and the plurality of training real-time scheduling costs can be used as labels to train the first neural network to obtain the active real-time scheduling cost model.

[0059] In a specific implementation, the plurality of training active real-time scheduling instructions and the corresponding plurality of training real-time scheduling costs, and the plurality of training scheduling instruction feasibility analysis results can be divided into a training set and a test set according to the needs of training and testing of the virtual power plant active real-time scheduling equivalent model. S Ttrain , D Train , , F Ctrain , F Atrain , and the training set is S Ttest , D Ttest , , F Atest , and the test set is S T is a training scheduling start period, D T is a training scheduling duration, is a training scheduling power schedule, F C is a training real-time scheduling cost, F A is a training scheduling instruction feasibility analysis result.

[0060] Then, the training data set S Ttrain , D Ttrain , is input to the deep neural network model, and F Ctrain is output, and the number of layers, the number of hidden nodes, the activation function, the optimization solver, and the size of the training data set of the model are determined by experiments. During the experiment, the size of the data set is gradually increased from small to large, and after the training is completed, the data set S Ttest , D Ttest , ,​​​​​F Ctest The test is performed until the model accuracy reaches the target accuracy requirement.

[0061] In S105, the second neural network is trained based on the multiple training active real-time scheduling instructions and the corresponding multiple training scheduling instruction feasibility analysis results, to obtain an active real-time scheduling instruction feasibility model; and the real-time scheduling cost and the scheduling instruction feasibility analysis result of the virtual power plant when executing the grid active real-time scheduling instruction are determined based on the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model.

[0062] In a specific implementation, the training data set S Ttrain , D Ttrain , is used as the input of the classification model, and F Atrain is used as the output, and a plurality of decision trees, support vector machines, nearest neighbors, and ensemble learning are used for training, and after the training, the data set S Ttest , D Ttest , , F Atest is used for testing, and the optimal classification model is selected.

[0063] The virtual power plant active real-time scheduling equivalent model construction method provided by the embodiment of the application comprises the following steps: obtaining multiple training active real-time scheduling instructions; obtaining an internal active real-time optimization scheduling model of the virtual power plant; determining corresponding multiple training real-time scheduling costs and multiple training scheduling instruction feasibility analysis results based on the internal active real-time optimization scheduling model of the virtual power plant and the multiple training active real-time scheduling instructions; training a first neural network based on the multiple training active real-time scheduling instructions and the corresponding multiple training real-time scheduling costs, to obtain an active real-time scheduling cost model; training a second neural network based on the multiple training active real-time scheduling instructions and the corresponding multiple training scheduling instruction feasibility analysis results, to obtain an active real-time scheduling instruction feasibility model; in this way, when the day-ahead generation plan of the virtual power plant has been formulated, the real-time scheduling cost and the scheduling instruction feasibility analysis result of the virtual power plant when executing the grid active real-time scheduling instruction can be determined based on the constructed active real-time scheduling cost model and the active real-time scheduling instruction feasibility model, so that for a given grid active real-time scheduling instruction, the active real-time scheduling cost of each virtual power plant can be quickly determined, and the active real-time scheduling of the virtual power plant by the grid is facilitated.

[0064] In an optional embodiment, in S102, the internal active real-time optimization scheduling model of the virtual power plant is obtained, which comprises the following steps:

[0065] obtaining a target function corresponding to the active real-time optimization scheduling model of the virtual power plant; the target function is related to a first operation cost function of the thermal power source in the virtual power plant within a scheduling period and a second operation cost function of the energy storage system power source within the scheduling period; determining a constraint condition corresponding to the target function; the constraint condition includes a first operation constraint condition corresponding to the thermal power source, a second operation constraint condition corresponding to the energy storage system power source, and a power balance constraint condition in the virtual power plant.

[0066] In the implementation, when the day-ahead generation plan of the virtual power plant has been formulated, the power purchased by the virtual power plant from the external power grid is a determined value, which can be equivalent to a negative load, and the wind power source, the photovoltaic power source and the hydroelectric power source of the virtual power plant are non-adjustable units, which can also be equivalent to a negative load. Therefore, in the active real-time optimization scheduling of the virtual power plant, only the adjustable units such as the micro gas turbine and the energy storage are considered, and the adjustable units respond to the active real-time scheduling instruction of the power grid. Therefore, the target function of the active real-time optimization scheduling model of the virtual power plant is expressed as follows:

[0067] ;

[0068] wherein, T is the number of time periods included in a scheduling period, C MT (t) is the operation cost of the thermal power source in the t period, that is, the operation cost of the micro gas turbine, C ES (t) represents the operation cost of the energy storage power source in the t period, that is, the operation cost of the energy storage unit, and the calculation formulas of the respective parts are as follows:

[0069] 1) Operation cost of the micro gas turbine

[0070] ;

[0071] wherein, C MT,m represents the unit power cost of the micro gas turbine unit m, N MT represents the number of micro gas turbine units, P MT,m (t) represents the output power of the micro gas turbine unit m in the t period, is the length of a time period, C MT,m represents the start-up cost of the micro gas turbine unit m, u su,m represents the start-up variable of the micro gas turbine unit m in the t period, which is 1 when starting up and 0 when not starting up.

[0072] In some embodiments, since it is in the case of having formulated virtual power plant day-ahead generation plan, in response to the grid active real-time dispatching instruction, the output power of the micro gas turbine unit at time period t outside the training scheduling duration can be a determined value.

[0073] 2) Operating cost of energy storage unit

[0074] ,

[0075] where N ES represents the number of energy storage units, C ES,e represents the discharge cost of the energy storage unit e when discharging, in units of yuan / kWh; represents the discharge power of the unit e, in units of kW.

[0076] Since the charging power of the energy storage unit is derived from other power sources, the electricity cost has been included in the production cost of other power sources, so the charging cost of the energy storage unit is not counted here. The power output of the energy storage system P ES,e (t) can be represented as (the value is positive indicating discharging, and the value is negative indicating charging):

[0077] .

[0078] In some embodiments, since it is in the case of having formulated virtual power plant day-ahead generation plan, in response to the grid active real-time dispatching instruction, the output power of the energy storage unit at time period t outside the training scheduling duration can be a determined value.

[0079] The first operating constraint condition corresponding to the thermal power source includes:

[0080] ① Power upper and lower limit constraint:

[0081] ,

[0082] ,

[0083] where , respectively represent the maximum output and minimum output of the unit m, in units of kW; u MT,m (t) is an operating variable, and is 1 when operating, and is 0 otherwise; β is a safety margin coefficient.

[0084] ② Ramp rate constraint

[0085] ,

[0086] ,

[0087] wherein , denote the up-regulation ramp rate limit and the down-regulation ramp rate limit of unit m, respectively, in units of kW / min.

[0088] ③ Minimum shutdown time constraint

[0089] ,

[0090] wherein is the minimum shutdown time of unit m, in units of time period; u sd,m (t) is the shutdown variable, which is 1 if the unit is shut down, and 0 otherwise.

[0091] ④ Continuous operation time constraint

[0092] Minimum continuous operation time constraint:

[0093] ,

[0094] wherein is the minimum continuous operation time of unit m, in units of time period. The second operation constraint condition corresponding to the energy storage system power source comprises:

[0095] ① Charge and discharge constraint

[0096] ,

[0097] ,

[0098] wherein , denote the maximum discharge power and the maximum charge power of unit e, respectively, , denote the minimum discharge power and the minimum charge power of unit e, respectively, in units of kW; , denote the discharge state and the charge state, respectively, and take the value of 1 to represent discharging / charging and take the value of 0 to represent not discharging / not charging, both satisfying the exclusivity, i.e. satisfying the following constraint:

[0099] .

[0100] ② SOC constraint

[0101] ,

[0102] wherein SOC e (t) denotes the SOC of unit e at time period t;SOC emax (t) 、 SOC emin (t) respectively represent the maximum and minimum allowable values of the SOC; SOC e (0) represents the initial capacity of the SOC; SOC e (t) represents the SOC capacity of the last period.

[0103] In some embodiments, the internal power balance constraint of the virtual power plant comprises:

[0104] The sum of the output power of the internal thermal power source of the virtual power plant, the output power of the energy storage system power source, and the training scheduling power corresponding to the training active real-time scheduling instruction, is equal to the total power of the internal load of the virtual power plant within the training scheduling duration.

[0105] In specific implementation, when the day-ahead generation plan of the virtual power plant has been formulated, the power purchased from the external power grid by the virtual power plant is a determined value, which can be equivalent to a negative load, and the wind power source, the photovoltaic power source, and the hydroelectric power source of the virtual power plant are non-adjustable units, which can also be equivalent to a negative load. Therefore, in the internal active real-time optimization scheduling of the virtual power plant, only the adjustable units such as micro gas turbines and energy storage systems are considered, and the adjustable units respond to the real-time scheduling instruction of the power grid. Therefore, the internal power balance constraint of the virtual power plant is:

[0106] ;

[0107] In the formula, N MT is the number of micro gas turbine units, P MT,m (t) represents the output power of the micro gas turbine unit m at the t period; N ES represents the number of energy storage units; P ES,e (t) represents the output power of the energy storage unit, which is positive when discharging and negative when charging; is the real-time scheduling power of the power grid at the t period, which is positive when increasing the power from the external power grid to the virtual power plant; is the total real-time net load power of the virtual power plant at the t period after excluding the non-adjustable units and the power purchased from the external power grid.

[0108] In the embodiment, since the virtual power plant is in the case of having formulated the day-ahead generation plan of the virtual power plant, and in response to the active real-time dispatching instruction of the power grid, the total power of the load in the virtual power plant in the t period is a determined value, therefore, is a determined value.

[0109] In the embodiment, the setting of the internal power balance constraint condition of the virtual power plant includes: the sum of the output power of the internal thermal power source of the virtual power plant, the output power of the energy storage system power source, and the training dispatch power corresponding to the training active real-time dispatching instruction, is equal to the total power of the internal load of the virtual power plant in the training dispatch duration period; in this way, only the adjustable units such as micro gas turbines and energy storages are considered, and the internal power balance constraint condition of the virtual power plant and the corresponding objective function can be simplified by responding to the active real-time dispatching instruction of the power grid through the adjustable units.

[0110] In the embodiment, by obtaining the objective function corresponding to the active real-time optimization dispatching model of the virtual power plant, and determining the constraint condition corresponding to the objective function, the active real-time optimization dispatching model of the virtual power plant can be quickly constructed, so that the active real-time optimization dispatching model of the virtual power plant considers not only the real-time dispatching cost, but also the operation constraint condition of the energy storage system power source and the thermal power source in the virtual power plant, and also considers the dispatching demand of meeting the active real-time dispatching instruction.

[0111] In an optional embodiment, the step S101 of obtaining a plurality of training active real-time dispatching instructions includes:

[0112] The dispatching start period range, the dispatching duration period range, and the dispatching power range corresponding to the training active real-time dispatching instruction are determined; and based on the dispatching start period range, the dispatching duration period range, and the dispatching power range, a plurality of training active real-time dispatching instructions are randomly generated by using a uniform distribution function.

[0113] In specific implementation, the training dispatch start period data set may include:

[0114] ;

[0115] In the formula, N sample is the size of the data set; represents the start period of the lth sample data, and is randomly generated by a uniform distribution function by considering the possible value range, that is:

[0116] ;

[0117] In the formula, is a uniform distribution law with a value interval of , and are the allowed minimum value and maximum value of the start period, respectively.

[0118] The training scheduling duration dataset can include:

[0119]

[0120] wherein, Tl denotes the scheduling duration of the lth sample data, and likewise, according to the possible value range thereof, a uniform distribution function is used for random generation, that is:

[0121]

[0122] wherein, Tl denotes the scheduling duration of the lth sample data, and likewise, according to the possible value range thereof, a uniform distribution function is used for random generation, that is:

[0123] The training scheduling power dataset can include:

[0124]

[0125] wherein, Pl denotes the active real-time scheduling power of the lth sample data, and likewise, according to the possible value range thereof, a uniform distribution function is used for random generation, that is:

[0126]

[0127] wherein, Pl denotes the active real-time scheduling power of the lth sample data, and likewise, according to the possible value range thereof, a uniform distribution function is used for random generation, that is: In the embodiment, by determining the scheduling start duration range, the scheduling duration range and the scheduling power range corresponding to the training active real-time scheduling instruction, and based on the scheduling start duration range, the scheduling duration range and the scheduling power range, a plurality of training active real-time scheduling instructions are randomly generated by using a uniform distribution function; in this way, a plurality of training active real-time scheduling instructions randomly distributed can be quickly generated.

[0128] In an optional embodiment, in step S103, based on the virtual power plant internal active real-time optimization scheduling model and the plurality of training active real-time scheduling instructions, a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results are determined, including:

[0129]

[0130] ​​​​​​​​Solving the active power real-time optimization scheduling model inside the virtual power plant based on each training active power real-time scheduling instruction to obtain a solution; the solution includes a first output power curve of the thermal power source inside the virtual power plant in the training scheduling duration, a second output power curve of the energy storage system power source in the training scheduling duration, and a training minimum real-time scheduling cost of the virtual power plant in a corresponding scheduling cycle; if the solution shows that the first output power curve, the second output power curve, and the training minimum real-time scheduling cost have a solution, the training minimum real-time scheduling cost is taken as the corresponding training real-time scheduling cost; and it is determined that the training active power scheduling instruction feasibility analysis result is feasible; if the solution shows that the first output power curve, the second output power curve, or the training minimum real-time scheduling cost has no solution, it is determined that the training active power scheduling instruction feasibility analysis result is infeasible.

[0131] In specific implementation, the objective function corresponding to the active power real-time optimization scheduling model inside the virtual power plant can be solved based on each training active power real-time scheduling instruction and based on the constraint condition corresponding to the active power real-time optimization scheduling model inside the virtual power plant. If the first output power curve of the thermal power source inside the virtual power plant in the training scheduling duration, the second output power curve of the energy storage system power source in the training scheduling duration, and the training minimum real-time scheduling cost of the virtual power plant in a corresponding scheduling cycle can be solved, the training minimum real-time scheduling cost is taken as the corresponding training real-time scheduling cost, and it is determined that the training active power scheduling instruction feasibility analysis result is feasible; if the first output power curve of the thermal power source inside the virtual power plant in the training scheduling duration, the second output power curve of the energy storage system power source in the training scheduling duration, or the training minimum real-time scheduling cost of the virtual power plant in a corresponding scheduling cycle cannot be solved, it is determined that the training active power scheduling instruction feasibility analysis result is infeasible.

[0132] In this embodiment, the multiple training real-time scheduling costs corresponding to the multiple training active power real-time scheduling instructions and the multiple training scheduling instruction feasibility analysis results can be quickly determined.

[0133] The embodiment of the application provides a virtual power plant active power scheduling method, as shown in Figure 2 The embodiment of the application provides a virtual power plant active power scheduling method, as shown in

[0134] S201, obtaining an active power scheduling instruction of a power grid to a virtual power plant, the active power scheduling instruction including a scheduling start period, a scheduling duration, and a scheduling power.

[0135] S202, based on the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model corresponding to each virtual power plant, respectively process the active scheduling instruction to obtain the real-time scheduling cost and the scheduling instruction feasibility analysis result when each virtual power plant executes the active scheduling instruction; the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model are constructed by using the virtual power plant active real-time scheduling equivalent model construction method in any of the above embodiments.

[0136] S203, based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, determine the target virtual power plant, and use the target virtual power plant to execute the active scheduling instruction.

[0137] The virtual power plant active scheduling method provided in the embodiments of the present application obtains the active scheduling instruction of the power grid to the virtual power plant, the active scheduling instruction including a scheduling start period, a scheduling duration period and a scheduling power; based on the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model corresponding to each virtual power plant, respectively process the active scheduling instruction to obtain the real-time scheduling cost and the scheduling instruction feasibility analysis result when each virtual power plant executes the active scheduling instruction; the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model are constructed by using the virtual power plant active real-time scheduling equivalent model construction method in any of the above embodiments; based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, determine the target virtual power plant, and use the target virtual power plant to execute the active scheduling instruction; in this way, in the case that the day-ahead generation plan of the virtual power plant has been made, the real-time scheduling cost and the scheduling instruction feasibility analysis result when the virtual power plant executes the active real-time scheduling instruction of the power grid can be determined by using the constructed active real-time scheduling cost model and the active real-time scheduling instruction feasibility model, so that for a given active real-time scheduling instruction of the power grid, the active real-time scheduling cost of each virtual power plant can be quickly determined, which is convenient for the active real-time scheduling of the virtual power plant by the power grid.

[0138] In an optional embodiment, in step S203, based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, determine the target virtual power plant, including:

[0139] Based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, determine the virtual power plant whose scheduling instruction feasibility analysis result is feasible and whose real-time scheduling cost is the smallest as the target virtual power plant.

[0140] In the embodiment, by determining the virtual power plant whose scheduling instruction feasibility analysis result is feasible and whose real-time scheduling cost is the smallest as the target virtual power plant, the power grid can make the active real-time scheduling cost of the whole network the smallest in the active real-time optimization scheduling, thereby saving the cost of the active real-time optimization scheduling of the power grid.

[0141] The embodiment of the present application also provides a virtual power plant active real-time scheduling equivalent model construction device, as shown in the figure, comprising: Figure 3

[0142] The first acquisition module 31 is used for acquiring a plurality of training active real-time scheduling instructions, wherein the training active real-time scheduling instruction comprises a training scheduling start period, a training scheduling duration and training scheduling power.

[0143] The second acquisition module 32 is used for acquiring a virtual power plant internal active real-time optimization scheduling model; the virtual power plant internal active real-time optimization scheduling model meets the requirements of the training active real-time scheduling instruction, and determines the planned output power curve of each type of distributed power source in the virtual power plant in a scheduling period, and takes the minimum real-time scheduling cost of the virtual power plant in a scheduling period as the target.

[0144] The first determination module 33 is used for determining a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results based on the virtual power plant internal active real-time optimization scheduling model and the plurality of training active real-time scheduling instructions.

[0145] The first training module 34 is used for training the first neural network based on the plurality of training active real-time scheduling instructions and the plurality of training real-time scheduling costs, and obtaining an active real-time scheduling cost model.

[0146] The second training module 35 is used for training the second neural network based on the plurality of training active real-time scheduling instructions and the plurality of training scheduling instruction feasibility analysis results, and obtaining an active real-time scheduling instruction feasibility model; and the real-time scheduling cost and the scheduling instruction feasibility analysis result of the virtual power plant when executing the grid active real-time scheduling instruction are determined based on the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model.

[0147] The virtual power plant active real-time scheduling equivalent model construction device provided by the embodiment of the present application can determine the real-time scheduling cost and the scheduling instruction feasibility analysis result of the virtual power plant when executing the grid active real-time scheduling instruction through the constructed active real-time scheduling cost model and the active real-time scheduling instruction feasibility model in the case that the day-ahead generation plan of the virtual power plant has been made, so that the active real-time scheduling cost of each virtual power plant can be quickly determined for the given grid active real-time scheduling instruction, and the active real-time scheduling of the virtual power plant by the grid is facilitated.

[0148] The embodiment of the present application also provides a virtual power plant active scheduling device, as shown in the figure, comprising: Figure 4

[0149] ​​The third obtaining module 41 is configured to obtain the active power dispatch instruction of the power grid to the virtual power plant, wherein the active power dispatch instruction comprises a dispatch start time period, a dispatch duration time period and a dispatch power.

[0150] The processing module 42 is configured to process the active power dispatch instruction based on the active power real-time dispatch cost model and the active power real-time dispatch instruction feasibility model corresponding to each virtual power plant, to obtain the real-time dispatch cost and the dispatch instruction feasibility analysis result when each virtual power plant executes the active power dispatch instruction; the active power real-time dispatch cost model and the active power real-time dispatch instruction feasibility model are constructed by using the virtual power plant active power real-time dispatch equivalent model construction method in any of the above embodiments.

[0151] The second determining module 43 is configured to determine the target virtual power plant based on the real-time dispatch cost and the dispatch instruction feasibility analysis result corresponding to each virtual power plant, and to use the target virtual power plant to execute the active power dispatch instruction.

[0152] The virtual power plant active power dispatch device provided by the embodiment of the present application can, in the case that the day-ahead generation plan of the virtual power plant has been made, determine the real-time dispatch cost and the dispatch instruction feasibility analysis result when the virtual power plant executes the active power real-time dispatch instruction of the power grid by using the constructed active power real-time dispatch cost model and the active power real-time dispatch instruction feasibility model, so that the active power real-time dispatch cost of each virtual power plant can be quickly determined for the given active power real-time dispatch instruction of the power grid, and the active power real-time dispatch of the virtual power plant by the power grid is facilitated.

[0153] According to the embodiments of the present application, the present application further provides an electronic device.

[0154] Figure 5 A schematic block diagram of an example electronic device that can be used to implement embodiments of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices, and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application described and / or claimed in this document.

[0155] As Figure 5As shown, the electronic device includes a computing unit 801 that can perform various appropriate actions and processes in accordance with a computer program stored in a read only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. Various programs and data required for operation of the electronic device can also be stored in the RAM 803. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other through a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0156] A plurality of components in the electronic device are connected to the I / O interface 805, including an input unit 806 such as a keyboard, a mouse, and the like, an output unit 807 such as various types of displays, a speaker, and the like, a storage unit 808 such as a magnetic disk, an optical disk, and the like, and a communication unit 809 such as a network card, a modem, a wireless communication transceiver, and the like. The communication unit 809 allows the electronic device to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0157] The computing unit 801 can be various general and / or special purpose processing components having processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, and the like. The computing unit 801 performs various methods and processes described above, such as the virtual power plant active real-time dispatch equivalent model construction method or the virtual power plant active dispatch method. For example, in some embodiments, the virtual power plant active real-time dispatch equivalent model construction method or the virtual power plant active dispatch method can be implemented as a computer software program that is tangibly embodied in a machine-readable medium, such as the storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the virtual power plant active real-time dispatch equivalent model construction method or the virtual power plant active dispatch method described above can be performed. Alternatively, in other embodiments, the computing unit 801 can be configured to perform the virtual power plant active real-time dispatch equivalent model construction method or the virtual power plant active dispatch method by any other appropriate means, such as by means of firmware.

[0158] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, specially designed application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0159] Program code for carrying out methods of the present application can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program code, when executed by the processor or controller, causes the functions / acts specified in the flowcharts and / or block diagrams to be implemented. The program code can execute entirely on a machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

[0160] The embodiment of the present application provides a computer program product or computer program, which comprises computer instructions stored in a computer readable storage medium. A processor of a computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the virtual power plant active scheduling equivalence model construction method or the virtual power plant active scheduling method provided in the embodiment of the present application.

[0161] The embodiment of the present application provides a computer readable storage medium storing executable instructions, wherein the executable instructions are stored, and when the executable instructions are executed by a processor, the processor executes the virtual power plant active scheduling equivalence model construction method or the virtual power plant active scheduling method provided in the embodiment of the present application.

[0162] In some embodiments, a computer-readable storage medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium can be a machine-readable signal medium or a machine-readable storage medium. A computer-readable storage medium can include but is not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a computer-readable storage medium can include a wired or wireless connection, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0163] In some embodiments, executable instructions can be in the form of programs, software, software modules, scripts, or code that is written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages; and that is deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0164] By way of example, executable instructions can, but need not, reside in a file system's files, can be stored in a part of a file that holds other programs or data, can be stored as a single file dedicated to the program or code, or can be stored in multiple files, such as files that store one or more modules, sub programs, or code portions.

[0165] By way of example, executable instructions can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a communication network.

[0166] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0167] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0168] The computer system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server is generally established by computer programs running on the respective computers and having a client-server relationship to each other. The servers can be cloud servers, servers of a distributed system, or servers combined with a blockchain.

[0169] It should be understood that various forms of flow shown above can be used, re-ordered, added to, or deleted from without departing from the technology disclosed in this application. For example, the steps recited in this application can be executed in parallel, executed in sequence, or executed in a different order, as long as the desired results of the technology disclosed in this application are achieved, and this application is not limited herein.

[0170] In addition, the terms "first", "second", etc., are used herein only to describe different instances, and do not imply or suggest relative importance or a number of the indicated technical features. Therefore, the features defined with "first", "second", etc., can explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "a plurality" is two or more, unless otherwise specifically limited.

[0171] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for constructing an equivalent model of real-time active power dispatching of a virtual power plant, characterized in that, The equivalent model of the virtual power plant active real-time scheduling is: in the case that the daily planning output power curve of each type of distributed power source of the virtual power plant and the power curve of the power purchase from the external power grid are known, the mapping model between the active real-time scheduling instruction of the next day and the internal real-time scheduling cost of the virtual power plant, which is used to provide the real-time scheduling cost corresponding to any time and any scheduling power of the next day for the active real-time scheduling instruction of the power grid, and the construction method comprises: Obtain a plurality of training active real-time scheduling instructions, the training active real-time scheduling instructions comprising a training scheduling start period, a training scheduling duration and a training scheduling power; Obtain an internal active real-time optimization scheduling model of the virtual power plant; the internal active real-time optimization scheduling model of the virtual power plant takes the minimum real-time scheduling cost of the virtual power plant in a scheduling period as the target while meeting the requirements of the training active real-time scheduling instruction and determining the planned output power curve of each type of distributed power source in the virtual power plant within a scheduling period; Based on the internal active real-time optimization scheduling model of the virtual power plant and a plurality of the training active real-time scheduling instructions, determine a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results, comprising: Based on each training active real-time scheduling instruction and the corresponding constraint condition of the internal active real-time optimization scheduling model of the virtual power plant, solve the objective function corresponding to the internal active real-time optimization scheduling model of the virtual power plant; if the first output power curve of the thermal power source in the virtual power plant within the training scheduling duration, the second output power curve of the energy storage system power source within the training scheduling duration and the training minimum real-time scheduling cost of the virtual power plant within a corresponding scheduling period can be solved, the training minimum real-time scheduling cost is taken as the corresponding training real-time scheduling cost, and the training active scheduling instruction feasibility analysis result is determined to be feasible; if the first output power curve of the thermal power source in the virtual power plant within the training scheduling duration, the second output power curve of the energy storage system power source within the training scheduling duration or the training minimum real-time scheduling cost of the virtual power plant within a corresponding scheduling period cannot be solved, the training active scheduling instruction feasibility analysis result is determined to be infeasible; Train the first neural network based on a plurality of the training active real-time scheduling instructions and a plurality of the corresponding training real-time scheduling costs to obtain an active real-time scheduling cost model; Train the second neural network based on a plurality of the training active real-time scheduling instructions and a plurality of the training scheduling instruction feasibility analysis results to obtain an active real-time scheduling instruction feasibility model; based on the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model, determine the real-time scheduling cost and the scheduling instruction feasibility analysis result of the virtual power plant when executing the active real-time scheduling instruction of the power grid.

2. The virtual power plant active real-time scheduling equivalence model construction method according to claim 1, characterized in that, Obtain an internal active real-time optimization scheduling model of the virtual power plant, comprising: obtain a target function corresponding to the active power real-time optimization scheduling model in the virtual power plant; the target function is related to a first operation cost function of the thermal power source in the virtual power plant within a scheduling period and a second operation cost function of the energy storage system power source within the scheduling period; determine a constraint condition corresponding to the target function; the constraint condition includes a first operation constraint condition corresponding to the thermal power source, a second operation constraint condition corresponding to the energy storage system power source, and a power balance constraint condition in the virtual power plant.

3. The virtual power plant active real-time scheduling equivalence model construction method according to claim 2, characterized in that, The power balance constraint condition in the virtual power plant includes: Within a training scheduling duration, the sum of the output power of the thermal power source, the output power of the energy storage system power source, and the training active power real-time scheduling instruction in the virtual power plant is equal to the total power of the load in the virtual power plant.

4. The virtual power plant active real-time scheduling equivalence model construction method according to claim 1, characterized in that, Obtaining a plurality of training active power real-time scheduling instructions includes: determining a scheduling start time range, a scheduling duration range, and a scheduling power range corresponding to the training active power real-time scheduling instruction; Based on the scheduling start time range, the scheduling duration range, and the scheduling power range, a plurality of training active power real-time scheduling instructions are randomly generated using a uniform distribution function.

5. The virtual power plant active real-time scheduling equivalence model construction method according to claim 1, characterized in that, Based on the active power real-time optimization scheduling model in the virtual power plant and the plurality of training active power real-time scheduling instructions, a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results are determined, including: Based on each training active power real-time scheduling instruction, the active power real-time optimization scheduling model in the virtual power plant is solved to obtain a solution; the solution includes a first output power curve of the thermal power source in the virtual power plant within a training scheduling duration, a second output power curve of the energy storage system power source within the training scheduling duration, and a training minimum real-time scheduling cost within a corresponding scheduling period of the virtual power plant; If the solution indicates that the first output power curve, the second output power curve, and the training minimum real-time scheduling cost have a solution, the training minimum real-time scheduling cost is taken as the corresponding training real-time scheduling cost; and the training active power scheduling instruction feasibility analysis result is determined to be feasible; If the solution indicates that the first output power curve, the second output power curve, or the training minimum real-time scheduling cost has no solution, the training active power scheduling instruction feasibility analysis result is determined to be infeasible.

6. A virtual power plant active scheduling method, characterized in that, including: Obtaining an active power scheduling instruction of a power grid to a virtual power plant, the active power scheduling instruction including a scheduling start time, a scheduling duration, and a scheduling power; Based on the active power real-time scheduling cost model and the active power real-time scheduling instruction feasibility model corresponding to each virtual power plant, the active power scheduling instruction is processed respectively to obtain the real-time scheduling cost and the scheduling instruction feasibility analysis result when each virtual power plant executes the active power scheduling instruction; the active power real-time scheduling cost model and the active power real-time scheduling instruction feasibility model are constructed by using the virtual power plant active power real-time scheduling equivalent model construction method in any one of claims 1-5. Based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, a target virtual power plant is determined, and the target virtual power plant is used to execute the active scheduling instruction.

7. The virtual power plant active scheduling method according to claim 6, characterized in that, Based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, a target virtual power plant is determined, and the target virtual power plant is used to execute the active scheduling instruction. Based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, a target virtual power plant is determined, and the target virtual power plant is used to execute the active scheduling instruction.

8. A virtual power plant active real-time dispatch equivalent model construction device, characterized in that, The virtual power plant active real-time scheduling equivalent model is a mapping model between the next day time period active real-time scheduling instruction and the virtual power plant internal real-time scheduling cost under the condition that the day-ahead planning output power curve of each type of distributed power supply of the virtual power plant and the power curve of the power purchased from the external power grid are known, and is used to provide the real-time scheduling cost corresponding to any time and any scheduling power of the next day for the grid active real-time scheduling instruction. The first obtaining module is configured to obtain a plurality of training active real-time scheduling instructions, wherein the training active real-time scheduling instruction comprises a training scheduling start period, a training scheduling duration and a training scheduling power. The second obtaining module is configured to obtain a virtual power plant internal active real-time optimization scheduling model, wherein the virtual power plant internal active real-time optimization scheduling model meets the requirement of the training active real-time scheduling instruction, and determines the planned output power curve of each type of distributed power supply in the virtual power plant within a scheduling period, and takes the minimum real-time scheduling cost of the virtual power plant within a scheduling period as a target. The first determining module is configured to determine a plurality of training real-time scheduling costs and a plurality of training scheduling instruction feasibility analysis results based on the virtual power plant internal active real-time optimization scheduling model and the plurality of training active real-time scheduling instructions, and is further configured to solve a target function corresponding to the virtual power plant internal active real-time optimization scheduling model based on each training active real-time scheduling instruction and a constraint condition corresponding to the virtual power plant internal active real-time optimization scheduling model; if the first output power curve of the thermal power supply within the training scheduling duration, the second output power curve of the energy storage system power supply within the training scheduling duration and the training minimum real-time scheduling cost of the virtual power plant within a corresponding scheduling period can be solved, the training minimum real-time scheduling cost is taken as the corresponding training real-time scheduling cost, and the training active scheduling instruction feasibility analysis result is determined to be feasible; if the first output power curve of the thermal power supply within the training scheduling duration, the second output power curve of the energy storage system power supply within the training scheduling duration or the training minimum real-time scheduling cost of the virtual power plant within a corresponding scheduling period cannot be solved, the training active scheduling instruction feasibility analysis result is determined to be infeasible. The first training module is configured to train the first neural network based on the plurality of training active real-time scheduling instructions and the plurality of corresponding training real-time scheduling costs, and obtain an active real-time scheduling cost model. The second training module is configured to train the second neural network based on the multiple training active real-time scheduling instructions and the corresponding multiple training scheduling instruction feasibility analysis results, and obtain an active real-time scheduling instruction feasibility model; and determine the real-time scheduling cost and the scheduling instruction feasibility analysis result of the virtual power plant when executing the grid active real-time scheduling instruction based on the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model.

9. A virtual power plant active scheduling apparatus, characterized by, The method comprises the following steps: The third obtaining module is configured to obtain an active scheduling instruction of the grid to the virtual power plant, wherein the active scheduling instruction comprises a scheduling start period, a scheduling duration period and a scheduling power; The processing module is configured to process the active scheduling instruction of each virtual power plant based on the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model corresponding to each virtual power plant, and obtain the real-time scheduling cost and the scheduling instruction feasibility analysis result of each virtual power plant when executing the active scheduling instruction; the active real-time scheduling cost model and the active real-time scheduling instruction feasibility model are constructed by using the virtual power plant active real-time scheduling equivalent model construction method in any one of claims 1 to 5; The second determining module is configured to determine a target virtual power plant based on the real-time scheduling cost and the scheduling instruction feasibility analysis result corresponding to each virtual power plant, and use the target virtual power plant to execute the active scheduling instruction.

10. An electronic device, comprising: The method comprises the following steps: At least one processor; and a memory connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the virtual power plant active real-time scheduling equivalent model construction method in any one of claims 1 to 5, or the virtual power plant active scheduling method in claim 6 or 7.

Images