Energy dispatching methods, devices and energy dispatching systems

Abstract

This application discloses an energy scheduling method, apparatus, and system, belonging to the field of energy scheduling technology. The energy scheduling method includes: acquiring the predicted working parameters and target data corresponding to the current sub-period and the remaining sub-periods within a target time period; sequentially inputting the predicted working parameters and target data corresponding to each sub-period into an energy scheduling model to obtain a sub-strategy corresponding to each sub-period; sequentially inputting the sub-strategy corresponding to each sub-period into a first simulation model to obtain the sub-execution result corresponding to each sub-period; adjusting the model parameters of the energy scheduling model based on the simulation results obtained from the sub-execution results of each sub-period within the target time period, and re-executing the steps of acquiring the predicted working parameters and target data corresponding to the current sub-period and the remaining sub-periods within the target time period until a preset stopping condition is reached; and selecting the target scheduling strategy corresponding to the target application object based on the simulation results.

Description

Technical Field

[0001] This application relates to the field of energy dispatching technology, and more specifically, to an energy dispatching method, apparatus and system. Background Technology

[0002] With the development of clean energy, the power system is gradually moving from the original "source follows load" model to an "integrated source-grid-load-storage" model. In related technologies, fixed dispatch strategies are mainly used to control and dispatch microgrids and equipment within energy stations, resulting in poor dispatch performance and difficulty in meeting user needs. Summary of the Invention

[0003] This application provides an energy dispatching method, apparatus, and system that are applicable to complex and ever-changing dispatching scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with dispatching strategies that meet their needs as much as possible, thereby improving the user experience.

[0004] In a first aspect, embodiments of this application provide an energy scheduling method, including:

[0005] Obtain the current sub-period and the prediction parameters and target data for each remaining sub-period within the target time period;

[0006] The prediction working parameters and target data corresponding to each sub-time period are sequentially input into the energy scheduling model to obtain the sub-strategy corresponding to each sub-time period.

[0007] The sub-strategies corresponding to the sub-time periods are sequentially input into the first simulation model to obtain the sub-execution results corresponding to each sub-time period.

[0008] When the simulation results are obtained based on the sub-execution results corresponding to each sub-period in the target time period, the model parameters of the energy scheduling model are adjusted, and the steps of obtaining the prediction working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period are re-executed until the preset stopping condition is reached.

[0009] The target scheduling strategy corresponding to the target application object is obtained by filtering based on the simulation results.

[0010] In the above technical solution, the energy scheduling model generates sub-strategies based on predicted short-term data, and the first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target scheduling strategy that better meets user needs. This approach is applicable to complex and ever-changing scheduling scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with scheduling strategies that meet their needs as much as possible, thereby improving the user experience.

[0011] In some embodiments, obtaining the prediction parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period includes:

[0012] Based on the prediction parameters corresponding to the previous sub-period of the current sub-period and the power prediction data in the target data, the prediction parameters corresponding to the current sub-period are obtained.

[0013] In the above technical solution, the predicted working parameters for the next sub-period are predicted by the predicted working parameters of the previous sub-period, which has high real-time performance and can solve the problem of not being able to obtain real-time working parameters for moments that have not yet occurred. This simulates the real scheduling environment, is suitable for complex and ever-changing scheduling environments, and improves the accuracy and real-time performance of the obtained scheduling strategy.

[0014] In some embodiments, obtaining the prediction parameters corresponding to the current sub-period based on the prediction parameters corresponding to the previous sub-period and the power prediction data in the target data includes:

[0015] The prediction parameters corresponding to the previous sub-period of the current sub-period and the power prediction data in the target data are input into the second simulation model to obtain the prediction parameters corresponding to the sub-period.

[0016] In the above technical solution, the real-time data of the next sub-period is obtained by simulating the real-time data of the previous sub-period. The scheduling strategy is predicted based on the real-time data of each sub-period obtained by simulation. It has high real-time performance, can simulate the real scheduling environment, is suitable for complex and ever-changing scheduling environments, and improves the accuracy and real-time performance of the obtained scheduling strategy.

[0017] In some embodiments, obtaining the prediction parameters corresponding to the current sub-period based on the prediction parameters corresponding to the previous sub-period and the power prediction data in the target data includes:

[0018] The second simulation model is adjusted based on the sub-execution result corresponding to the previous sub-period of the current sub-period.

[0019] The prediction parameters corresponding to the previous sub-period and the power prediction data in the target data are input into the adjusted second simulation model to obtain the prediction parameters corresponding to the current sub-period.

[0020] In the above technical solution, the second simulation model of the current sub-period is dynamically updated by the sub-execution result corresponding to the previous sub-period. This can more realistically simulate the dynamic changes in the working status of each device during the actual scheduling process of the target application object, improve the accuracy and realism of the simulation results, and better adapt to the requirements of "real-time or near-real-time updates" brought about by various complex factors such as seasonal changes, weather changes, and changes in user energy consumption scenarios on the scheduling strategy.

[0021] In some embodiments, the step of sequentially inputting the prediction working parameters and target data corresponding to each of the sub-time periods into the energy scheduling model to obtain the sub-strategy corresponding to each of the sub-time periods includes:

[0022] For each sub-period, the sub-period is taken as the current sub-period, and the first simulation model is adjusted according to the sub-execution result corresponding to the current sub-period;

[0023] The sub-strategy corresponding to the next sub-period of the current sub-period is input into the adjusted first simulation model to obtain the sub-execution result corresponding to the next sub-period of the current sub-period.

[0024] In the above technical solution, the first simulation model of the current sub-period is dynamically updated by the sub-execution result corresponding to the previous sub-period. This can more realistically simulate the dynamic changes in the working status of each device during the actual scheduling process of the target application object, improve the accuracy and realism of the simulation results, and better adapt to the requirements of "real-time or near-real-time updates" brought about by various complex factors such as seasonal changes, weather changes, and changes in user energy consumption scenarios on the scheduling strategy.

[0025] In some embodiments, the step of sequentially inputting the prediction working parameters and target data corresponding to each of the sub-time periods into the energy scheduling model to obtain the sub-strategy corresponding to each of the sub-time periods includes:

[0026] The sub-strategy corresponding to the sub-time period is input into the first simulation model so that at least one target device model in the first simulation model can execute the sub-strategy to obtain the sub-execution result corresponding to the sub-time period. The sub-execution result includes the simulation working parameters of the target device model and the device model associated with the target device model after executing the sub-execution result. The simulation working parameters are used to adjust the first simulation model so that the adjusted first simulation model can simulate and deduce the sub-execution result corresponding to the next sub-time period of the sub-time period.

[0027] In the above technical solution, by constructing a simulation model to execute the sub-strategies corresponding to each sub-period in chronological order, the simulation model can be dynamically updated based on the sub-execution results of the previous sub-period. This drives the updated simulation model to execute the sub-strategy corresponding to the current sub-period, which has a high adaptive adjustment capability. This allows for a more realistic simulation of the target application object in actual operation, including the linkage between various devices and the complex and ever-changing environmental conditions, thereby improving the accuracy and real-time performance of the simulation results and meeting user needs.

[0028] In some embodiments, the step of sequentially inputting the prediction working parameters and target data corresponding to each of the sub-time periods into the energy scheduling model to obtain the sub-strategy corresponding to each of the sub-time periods includes:

[0029] The prediction parameters corresponding to the previous sub-period of each sub-period are simulated and deduced using the second simulation model until the prediction parameters corresponding to each sub-period included in the target period are obtained.

[0030] The energy scheduling model processes the prediction parameters and target data for each sub-period within the target time period to obtain the scheduling strategy corresponding to the target time period. The scheduling strategy includes sub-strategies corresponding to each sub-period.

[0031] In the above technical solution, the predicted working parameters of the current sub-period are predicted from the predicted working parameters of the previous sub-period through the second simulation model, and are used as the real-time working parameters of the current sub-period. The sub-strategy of the current sub-period is predicted by the energy scheduling model based on the real-time working parameters of the current sub-period. This can quickly and efficiently simulate and deduce the working status of different sub-periods within the target period, and has high real-time performance. It can adapt to the requirements of "real-time or near-real-time updates" brought about by various complex factors such as seasonal changes, weather changes, and changes in user energy consumption scenarios.

[0032] In some embodiments, the step of adjusting the model parameters of the energy scheduling model and re-executing the step of obtaining the prediction working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period, when the simulation results are obtained based on the sub-execution results corresponding to each sub-period in the target time period, includes:

[0033] The following operations are performed sequentially according to the order of the sub-time periods:

[0034] S1. Initialize the current sub-time period;

[0035] S2. Obtain the prediction working parameters and target data corresponding to the current sub-time period;

[0036] S3. Input the prediction working parameters and target data corresponding to the current sub-period into the energy scheduling model to obtain the sub-strategy corresponding to the current sub-period;

[0037] S4. Input the sub-strategy corresponding to the current sub-period into the first simulation model to obtain the sub-execution result corresponding to the sub-strategy corresponding to the current sub-period;

[0038] S5. If the current sub-period is not the last sub-period of the target period, then return to execute S1;

[0039] S6. If the current sub-time period is the last sub-time period of the target time period, then terminate the operation.

[0040] In the above technical solution, by constructing a dual-loop framework of small loop and large loop of simulation data, it is possible to obtain simulation results that are closer to the current user needs, and then select target scheduling strategies for reference or execution of target application objects. This can meet the needs of real-time or near real-time operation and make the scheduling results more in line with the current needs of users.

[0041] In some embodiments, when the simulation results are obtained based on the sub-execution results corresponding to each sub-time period in the target time period, the model parameters of the energy scheduling model are adjusted, and the steps of obtaining the prediction working parameters and target data corresponding to the current sub-time period and the remaining sub-time periods in the target time period are re-executed until a preset stopping condition is reached, including:

[0042] If the number of adjustments is less than or equal to the first threshold, the step of obtaining the current sub-period and the prediction working parameters and target data corresponding to each remaining sub-period in the target period is re-executed.

[0043] If the number of adjustments exceeds the first threshold, the step of obtaining the prediction working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period shall be stopped.

[0044] In the above technical solution, by constructing a dual-loop framework of small loop and large loop of simulation data, it is possible to obtain simulation results that are closer to the current user needs, and then select target scheduling strategies for reference or execution of target application objects. This can meet the needs of real-time or near real-time operation and make the scheduling results more in line with the current needs of users.

[0045] In some embodiments, adjusting the model parameters of the energy dispatch model includes at least one of the following methods:

[0046] The model parameters of the energy dispatch model are adjusted with the goal of obtaining the optimal value of the simulation results;

[0047] or,

[0048] Randomly select some parameters of the energy dispatch model to determine the parameters to be adjusted;

[0049] Adjust the parameter to be adjusted so that the adjusted parameter is different from the parameter after each previous adjustment;

[0050] or,

[0051] The model parameters of the energy scheduling model are adjusted using a tuner.

[0052] or,

[0053] The model parameters of the energy scheduling model are adjusted using a loss function; the loss function is constructed based on the simulation results and the target simulation results.

[0054] In the above technical solution, by fine-tuning the model parameters to obtain a variety of different energy scheduling algorithms, multiple different scheduling strategies can be quickly obtained for selection. Combined with the simulation model, the scheduling strategies are simulated and deduced to obtain a scheduling strategy that is as close as possible to the user's needs. It has high flexibility and is simple to operate and easy to implement.

[0055] In some embodiments, the step of obtaining the target scheduling strategy corresponding to the target application object based on the simulation results includes:

[0056] Based on the simulation results, calculate the evaluation score corresponding to the simulation results from at least one dimension;

[0057] The scheduling strategy corresponding to the simulation result with the highest evaluation score is determined as the target scheduling strategy.

[0058] In the above technical solution, the simulation execution of the scheduling strategy is performed to obtain the simulation results corresponding to the scheduling strategy. Based on the simulation inference capability, the simulation results are evaluated according to user needs and one or more dimensions. With the goal of selecting the one that is closest to the user's current needs, various prediction algorithms and scheduling algorithms in the energy scheduling model are trained and optimized. This results in a scheduling strategy and model parameters that are closest to the user's current needs under the influence of multiple factors such as power generation capacity, power load, grid supply capacity, and electricity price during the target time period. This enables real-time or near-real-time high-frequency iteration, self-iteration, and self-optimization, adapting to complex and ever-changing scheduling environments and meeting user needs.

[0059] In some embodiments, the at least one dimension includes the total revenue corresponding to the target time period.

[0060] In the above technical solutions, optimizing the total revenue is the goal, which can effectively guarantee user operational benefits.

[0061] In some embodiments, before obtaining the prediction working parameters and target data corresponding to the current sub-time period and the remaining sub-time periods in the target time period, the method further includes:

[0062] The first simulation model and the second simulation model are constructed based on at least one of the following: the device information of the target application object, the local control strategy corresponding to the device included in the target application object, the power topology information of the target application object, and the overall coordination control strategy of the target application object.

[0063] In the above technical solution, the target application object is simulated and modeled based on its actual characteristics. This yields a simulation model that simulates the local control strategies of each device within the target application object and the overall coordinated control strategy of the target application object. This provides real-time and near-real-time simulation capabilities, enabling the energy dispatch algorithm to track changes in the target application object's power generation, power consumption, energy storage, and grid power supply in real-time and near-real-time, as well as changes in information such as grid connection prices and electricity prices. This improves the optimization effect of the energy dispatch algorithm, allowing it to generate dispatch strategies that are closer to the user's current needs based on the real-time changes of the target application object, thus meeting the user's requirements.

[0064] In some embodiments, before obtaining the prediction working parameters and target data corresponding to the current sub-time period and the remaining sub-time periods in the target time period, the method further includes:

[0065] A training sample set is constructed using the sample operating parameters of the target application object corresponding to multiple sample sub-time periods as samples and the sample sub-strategies derived based on the sample operating parameters as labels. The sample operating parameters are obtained by simulation using a second simulation model based on the sample operating parameters and sample power prediction data corresponding to the previous sample sub-time period. The sample operating parameters are used to characterize the real-time operating parameters of the target application object in each of the sample sub-time periods.

[0066] The energy scheduling model is constructed based on the training sample set.

[0067] In the above technical solution, during the construction and training of the energy scheduling model, simulation data and sample scheduling strategies derived from the simulation data and simulation model are used to train the energy scheduling model, so that the energy scheduling model can better adapt to the simulation data, improve the accuracy and precision of the energy scheduling model in the simulation scenario, and thus improve the accuracy of the final scheduling strategy.

[0068] In some embodiments, after obtaining the target scheduling strategy corresponding to the target application object based on the simulation results, the method further includes:

[0069] The model parameters corresponding to the target scheduling strategy are determined as the actual model parameters of the energy scheduling model during the target time period.

[0070] In the above technical solution, by determining the model parameters corresponding to the target scheduling strategy as the actual model parameters of the energy scheduling model in the target time period, the energy scheduling model can be updated in the short term with the goal of optimizing the simulation results. This makes the prediction results of the energy scheduling model in the short term more in line with the user's short-term needs, and the model parameters of the energy scheduling model can be dynamically adjusted according to changes in user needs, which has high flexibility.

[0071] Secondly, embodiments of this application provide an energy scheduling method applied to a simulation model, the simulation model including a first simulation model; the method includes:

[0072] When receiving the sub-policy corresponding to a sub-time period within the target time period sent by the energy scheduling model, the first simulation model performs simulation deduction based on the sub-policy to obtain the sub-execution result corresponding to the sub-time period; wherein...

[0073] The sub-execution results are used to determine the simulation results corresponding to the target time period, which includes multiple sub-time periods. The simulation results are used to adjust the model parameters of the energy scheduling model to obtain the target scheduling strategy corresponding to the target application object.

[0074] The sub-strategy corresponding to the sub-period is predicted by the energy scheduling model based on the prediction working parameters and target data corresponding to the sub-period.

[0075] Thirdly, embodiments of this application provide an energy scheduling method applied to an energy scheduling model, the method comprising:

[0076] Upon receiving the prediction parameters corresponding to each sub-period within the target time period, the sub-strategy corresponding to each sub-period is predicted sequentially based on the prediction parameters corresponding to each sub-period and the target data; the target time period includes multiple sub-periods.

[0077] The sub-strategies corresponding to the sub-time periods are sequentially sent to the first simulation model for the first simulation model to execute the sub-strategies, obtain the sub-execution results corresponding to each sub-time period, and obtain the simulation results corresponding to the target time period based on each sub-execution result;

[0078] After obtaining the simulation results corresponding to the target time period and adjusting the model parameters of the energy scheduling model, the step of predicting the sub-strategy corresponding to each sub-time period based on the prediction working parameters corresponding to each sub-time period and the target data is re-executed when the prediction working parameters corresponding to each sub-time period in the target time period are received, until the preset stopping condition is reached; so as to filter the target scheduling strategy corresponding to the target application object according to the obtained simulation results.

[0079] Fourthly, embodiments of this application provide an energy scheduling method applied to an energy scheduling system, the energy scheduling system including a controller, an energy scheduling model, and a simulation model deployed on a simulation service cluster, wherein the simulation model and the energy scheduling model are connected; the method includes:

[0080] The energy scheduling model sequentially predicts the prediction working parameters and target data corresponding to each sub-period in the target time period to obtain the sub-strategy corresponding to each sub-period; the target time period includes multiple sub-periods.

[0081] In the simulation model, the first simulation model sequentially simulates and deduces each of the sub-strategies to obtain the sub-execution results corresponding to each of the sub-time periods;

[0082] When simulation results are obtained based on the sub-execution results corresponding to each sub-time period in the target time period, the controller adjusts the model parameters of the energy scheduling model;

[0083] The adjusted energy scheduling model re-executes the steps of predicting the prediction working parameters and target data corresponding to each sub-time period in the target time period in turn to obtain the sub-strategy corresponding to each sub-time period, until the preset stopping condition is reached;

[0084] The controller selects the target scheduling strategy corresponding to the target application object based on the simulation results.

[0085] Fifthly, embodiments of this application provide an energy dispatching device, comprising:

[0086] The first processing module is used to obtain the current sub-period and the prediction working parameters and target data corresponding to each remaining sub-period in the target time period;

[0087] The second processing module is used to sequentially input the prediction working parameters and target data corresponding to each sub-time period into the energy scheduling model to obtain the sub-strategy corresponding to each sub-time period.

[0088] The third processing module is used to sequentially input the sub-strategies corresponding to the sub-time periods into the first simulation model to obtain the sub-execution results corresponding to each sub-time period.

[0089] The fourth processing module is used to adjust the model parameters of the energy scheduling model based on the simulation results obtained from the sub-execution results corresponding to each sub-time period in the target time period, and to re-execute the step of obtaining the prediction working parameters and target data corresponding to the current sub-time period and the remaining sub-time periods in the target time period until the preset stopping condition is reached.

[0090] The fifth processing module is used to filter and obtain the target scheduling strategy corresponding to the target application object based on the simulation results.

[0091] Sixthly, embodiments of this application provide an energy dispatching device applied to a simulation model, the simulation model including a first simulation model; the device includes:

[0092] The sixth processing module is used to, upon receiving the sub-strategy corresponding to the sub-period in the target time period sent by the energy scheduling model, enable the first simulation model to perform simulation deduction based on the sub-strategy to obtain the sub-execution result corresponding to the sub-period; wherein, the sub-execution result is used to determine the simulation result corresponding to the target time period, the target time period includes multiple sub-periods, and the simulation result is used to adjust the model parameters of the energy scheduling model to obtain the target scheduling strategy corresponding to the target application object;

[0093] The sub-strategy corresponding to the sub-period is predicted by the energy scheduling model based on the prediction working parameters and target data corresponding to the sub-period.

[0094] Seventhly, embodiments of this application provide an energy scheduling device applied to an energy scheduling model, the device comprising:

[0095] The seventh processing module is used to enable the energy scheduling model to predict the sub-strategy corresponding to each sub-time period based on the prediction working parameters corresponding to each sub-time period and the target data when it receives the prediction working parameters corresponding to each sub-time period in the target time period; the target time period includes multiple sub-time periods;

[0096] The eighth processing module is used to enable the energy scheduling model to send the sub-strategies corresponding to the sub-time periods to the first simulation model in sequence, so that the first simulation model can execute the sub-strategies, obtain the sub-execution results corresponding to each sub-time period, and obtain the simulation results corresponding to the target time period based on each sub-execution result;

[0097] The ninth processing module is used to, after obtaining the simulation results corresponding to the target time period and adjusting the model parameters of the energy scheduling model, cause the energy scheduling model to re-execute the step of predicting the sub-strategies corresponding to each sub-time period based on the prediction working parameters corresponding to each sub-time period and the target data, until a preset stopping condition is reached; so as to filter the target scheduling strategy corresponding to the target application object according to the obtained simulation results.

[0098] Eighthly, embodiments of this application provide an energy scheduling device applied to an energy scheduling system, the energy scheduling system including a controller, an energy scheduling model, and a simulation model deployed on a simulation service cluster, the simulation model and the energy scheduling model being connected; the device includes:

[0099] The tenth processing module is used to enable the energy scheduling model to sequentially predict the prediction working parameters and target data corresponding to each sub-period in the target time period, so as to obtain the sub-strategy corresponding to each sub-period; the target time period includes multiple sub-periods;

[0100] The eleventh processing module is used to enable the first simulation model in the simulation model to perform simulation deduction on each of the sub-strategies in sequence, and to obtain the sub-execution results corresponding to each of the sub-time periods;

[0101] The twelfth processing module is used to enable the controller to adjust the model parameters of the energy scheduling model when the simulation results are obtained based on the sub-execution results corresponding to each sub-time period in the target time period.

[0102] The thirteenth processing module is used to enable the adjusted energy scheduling model to re-execute the steps of predicting the prediction working parameters and target data corresponding to each sub-period in the target time period in turn, and obtaining the sub-strategy corresponding to each sub-period, until the preset stopping condition is reached.

[0103] The fourteenth processing module is used to enable the controller to filter and obtain the target scheduling strategy corresponding to the target application object based on the simulation results.

[0104] Ninthly, embodiments of this application provide an energy dispatching system, including:

[0105] Controller;

[0106] A simulation model is deployed on a simulation service cluster. The controller is connected to the simulation service cluster and controls the target computing node in the simulation service cluster to enable the model in the simulation model corresponding to the target computing node to perform simulation and deduction.

[0107] An energy scheduling model is connected to both the simulation model and the controller.

[0108] The controller is configured to execute the energy scheduling method as described in the first aspect.

[0109] In a tenth aspect, this application provides an electronic device including 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 energy scheduling method as described in the first, second, or third aspect above.

[0110] In one aspect, this application provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the energy scheduling method as described in the first, second, or third aspect above.

[0111] In a twelfth aspect, this application provides a computer program product, including a computer program that, when executed by a processor, implements the energy scheduling method as described in the first, second, or third aspect above. Attached Figure Description

[0112] Figure 1 One of the schematic flowcharts of an energy scheduling method provided in some embodiments of this application;

[0113] Figure 2 A second schematic flowchart illustrating an energy scheduling method provided in some embodiments of this application;

[0114] Figure 3 The third schematic flowchart illustrates the energy scheduling method provided in some embodiments of this application;

[0115] Figure 4 Fourth of a series of schematic flowcharts illustrating the energy scheduling method provided in some embodiments of this application;

[0116] Figure 5 Fifth of a series of schematic flowcharts illustrating the energy scheduling method provided in some embodiments of this application;

[0117] Figure 6 Sixth schematic flowchart of an energy scheduling method provided in some embodiments of this application;

[0118] Figure 7 Seventh schematic flowchart of an energy scheduling method provided in some embodiments of this application;

[0119] Figure 8 This is a schematic diagram illustrating the effect of the energy scheduling method provided in some embodiments of this application;

[0120] Figure 9Eighth schematic flowchart of an energy scheduling method provided in some embodiments of this application;

[0121] Figure 10 A flowchart illustrating one of the energy scheduling methods provided in some embodiments of this application;

[0122] Figure 11 The tenth is a flowchart illustrating an energy scheduling method provided in some embodiments of this application;

[0123] Figure 12 This is one of the structural schematic diagrams of an energy dispatching device provided in some embodiments of this application;

[0124] Figure 13 This is the second schematic diagram of the structure of an energy dispatching device provided in some embodiments of this application;

[0125] Figure 14 This is the third schematic diagram of the structure of the energy dispatching device provided in some embodiments of this application;

[0126] Figure 15 Fourth schematic diagram of the structure of the energy dispatching device provided in some embodiments of this application;

[0127] Figure 16 The diagram shows the structure of an electronic device provided in some embodiments of this application. Detailed Implementation

[0128] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0129] Unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used in the description of this application is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms "comprising" and "having," and any variations thereof, in the description, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the description, claims, or accompanying drawings of this application are used to distinguish different objects, not to describe a specific order or hierarchy.

[0130] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application can be combined with other embodiments.

[0131] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "attachment" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0132] The energy scheduling method, energy scheduling device, electronic device, and readable storage medium provided in this application will be described in detail below with reference to the accompanying drawings and through specific embodiments and application scenarios.

[0133] The energy scheduling method can be applied to the terminal, and can be executed by the hardware or software in the terminal.

[0134] The energy scheduling method provided in this application embodiment can be executed by an electronic device or a functional module or entity in an electronic device that can implement the energy scheduling method. The electronic devices mentioned in this application embodiment include, but are not limited to, mobile phones, tablet computers and computers.

[0135] The inventors discovered that most current industry solutions lack the near real-time performance required for integrated energy storage and grid-load systems. Furthermore, fixed strategies are employed for charging and discharging energy storage devices within microgrids / energy stations, for on-site consumption or grid connection of photovoltaic power generation, and for adjusting flexible loads. While some industry solutions utilize AI-generated dynamic scheduling strategies, they cannot perform frequent real-time or near real-time iterations and optimizations of AI scheduling algorithms / models. This makes it difficult to adapt to the demands of seasonal changes, weather variations, and changes in user energy consumption scenarios, which necessitate real-time or near real-time updates to scheduling strategies. Consequently, current industry solutions often fail to provide users with consistently optimal energy scheduling strategies, thus failing to meet user needs.

[0136] Based on the above considerations, in order to solve the problem that energy scheduling strategies lack real-time or near-real-time performance and thus fail to meet user needs, the inventors, after in-depth research, designed an energy scheduling method, including: obtaining the predicted working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period; sequentially inputting the predicted working parameters and target data corresponding to each sub-period into the energy scheduling model to obtain the sub-strategy corresponding to each sub-period; sequentially inputting the sub-strategy corresponding to each sub-period into the first simulation model to obtain the sub-execution result corresponding to each sub-period; adjusting the model parameters of the energy scheduling model based on the simulation results obtained from the sub-execution results corresponding to each sub-period in the target time period, and re-executing the steps of obtaining the predicted working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period until a preset stopping condition is reached; and selecting the target scheduling strategy corresponding to the target application object based on the simulation results.

[0137] In this energy dispatching method, an energy dispatching model predicts sub-strategies for each sub-period based on the predicted operating parameters and target data. A pre-built first simulation model corresponding to the target application executes each sub-strategy, simulating the entire target period to obtain simulation results. By adjusting the model parameters of the energy dispatching model and performing multiple repeated simulations, multiple different dispatching strategies corresponding to the target period and different simulation results obtained after the first simulation model executes each dispatching strategy can be obtained. Based on the simulation results, a target dispatching strategy is selected from multiple different dispatching strategies for the target application to use as a reference or for execution. This method comprehensively considers various factors such as the real-time changes in the power generation capacity, electricity load, grid supply capacity and price, and energy storage capacity of the target application. It adapts to the requirement of "real-time or near-real-time updates" brought about by various complex factors such as seasonal changes, weather changes, and changes in user energy consumption scenarios, continuously providing users with dispatching strategies that meet their needs as much as possible, thereby improving the user experience.

[0138] like Figure 1 As shown, the energy scheduling method includes steps 110, 120, 130, 140 and 150.

[0139] Step 110: Obtain the prediction parameters and target data prediction parameters for the current sub-period and the remaining sub-periods in the target time period;

[0140] In this step, the target time period includes multiple sub-time periods. The target time period can be a day or half a day, etc., and a sub-time period is a shorter period of time. The duration of a sub-time period can be user-defined, such as a sub-time period corresponding to 5 minutes, 1 minute, 30 seconds, or other durations.

[0141] The current sub-time period is determined according to the time sequence of each sub-time period. It is the sub-time period corresponding to the current simulation round. The remaining sub-time periods are the sub-time periods in the target time period that have not yet entered the corresponding simulation round. The sub-time period before the current sub-time period is the sub-time period that has completed the simulation round.

[0142] The predicted working parameters are high-frequency working parameters used to calculate the scheduling execution strategy of the target application object in the current sub-period of the target time period. These parameters include the predicted working parameters of one or more devices in the target application object in the current sub-period. These predicted working parameters can be parameters predicted by the simulation system.

[0143] The real-time operating parameters corresponding to the first sub-period of the target period can be parameters predicted by the simulation system or actual parameters obtained.

[0144] The target application object is the object that needs to be energy dispatched, which may include, but is not limited to, microgrids and energy stations. In actual application, it can be adaptively selected according to actual needs. In some embodiments, the target application object can be one or more application objects, and each application object is dispatched independently.

[0145] The target application can be applied to a variety of different scenarios, such as microgrids in factory environments, building environments, or other environments.

[0146] In some embodiments, the target application may include multiple devices such as energy storage devices, power generation devices, and electrical devices. Among them, power generation devices may include traditional power generation devices, photovoltaic power generation devices, wind power generation devices, tidal power generation devices, and other new energy power generation devices, etc., which are not limited herein.

[0147] like Figure 2 As shown, in some embodiments, the predictive operating parameters may include at least one of the following: power generation equipment, power consumption equipment, power storage equipment, state of charge of energy storage equipment, and grid incoming power for each sub-period.

[0148] Figure 5 Example of predictive operating parameters under one condition, including: real-time switch status, power load, incoming electrical data, photovoltaic power generation, and SOX of the energy storage station.

[0149] Target data may include historical operating parameters of the target application object, environmental data of the area where the target application object is located, meteorological data, electricity price information, predicted power generation information, and predicted electricity load, as well as other energy dispatch-related information. The time step of the target data can be consistent with the sub-time period.

[0150] In some embodiments, the target data may include one or more of the following: power forecast data, historical data, and electricity price data corresponding to each sub-period within the target period.

[0151] In some embodiments, historical data may include at least one of the following: historical meteorological data, historical power generation, historical load, and historical grid incoming power for the area where the target application is located; in some embodiments, the step size of the historical data may be consistent with the sub-time period.

[0152] Electricity price data refers to the electricity price data used to characterize each sub-period within the target period, such as time-of-use electricity price data and capacity electricity price data. It can be understood that the electricity price data corresponding to different sub-periods within the same target period may be dynamically changing.

[0153] The power forecast data consists of power generation and power consumption data for each sub-period within the target period, predicted based on historical operating parameters.

[0154] Continue to refer to Figure 2 In some embodiments, power forecast data may include at least one of the predicted power generation and predicted load for each sub-period, such as the predicted photovoltaic power generation data and the predicted load for each sub-period within the target period; in actual execution, power forecast data can be obtained by prediction through a pre-trained network model or a time series model.

[0155] like Figure 5 As shown, in some embodiments, the predicted load can be obtained based on historical data (such as historical meteorological data, historical electricity load, etc.), predicted operating parameters (such as predicted electricity load), and real-time meteorological data (such as weather forecasts and real-time weather, etc.).

[0156] Subsequent simulations and predictions of energy dispatch strategies based on electricity load can make the generated dispatch strategies more in line with users' energy consumption habits, and improve the adaptability and intelligence level of energy dispatch algorithms when facing different users.

[0157] In some embodiments, photovoltaic power generation forecast data can be obtained based on historical data (such as photovoltaic power generation capacity) and real-time meteorological data (such as weather forecasts and real-time weather).

[0158] Of course, in other embodiments, the target data may also include cost data, energy consumption data, and carbon emission data of the target application object during the target time period, which can be dynamically adjusted according to actual business needs.

[0159] Step 120: Input the prediction working parameters and target data corresponding to each sub-period into the energy scheduling model in sequence to obtain the sub-strategy corresponding to each sub-period;

[0160] In this step, the sub-strategy is the energy scheduling execution strategy corresponding to the target application object within the sub-period, which is used to control the operating status of the corresponding equipment in the target application object, including but not limited to controlling the switching status of each energy storage device, the charging and discharging power of each energy storage device, the charging and discharging power of charging equipment (such as charging piles), the power limit of the power generation equipment, the full power of the power generation equipment, the power consumption of flexible loads, and the working status of other equipment.

[0161] Understandably, different sub-strategies will correspond to different target application objects, depending on the differences in the devices they include. For example, for target application objects in a factory scenario, the sub-strategies will also include strategies for controlling the working status of each production device in the factory. Similarly, for target application objects in a building scenario, the sub-strategies will also include strategies for controlling the working status of elevator equipment in the building.

[0162] The energy scheduling model is a model used to predict scheduling strategies. The energy scheduling model is equipped with an energy scheduling algorithm. The input features of the energy scheduling model are the predicted working parameters and target data corresponding to the current sub-period, and the output features are the sub-strategies corresponding to the current sub-period. In actual execution, the energy scheduling model can be set to match the target application object.

[0163] The energy scheduling model can be an artificial intelligence model or a neural network model, etc., and this application does not limit it. The energy scheduling model can be trained based on simulation data and scheduling strategies derived from simulation data. The training method of the energy scheduling model will be described below, and will not be elaborated here.

[0164] The above embodiment will be explained using the target time period as an example, which includes multiple sub-time periods such as A, B, C, and D in sequential order. Here, t0, t1, t2, and t3 are the start times of time periods A, B, C, and D, respectively.

[0165] Continue to refer to Figure 2 In actual execution, the predicted working parameters and target data corresponding to time t0 can be input into the energy scheduling model, and the energy scheduling model can predict the sub-strategy corresponding to time period A; the predicted working parameters and target data corresponding to time t1 can be input into the energy scheduling model, and the energy scheduling model can predict the sub-strategy corresponding to time period B. This process is repeated until the sub-strategy corresponding to the last sub-time period is obtained.

[0166] Each sub-strategy is arranged in chronological order to obtain the scheduling strategy corresponding to the target time period.

[0167] In some embodiments, the output features of the energy scheduling model may include sub-strategies corresponding to each sub-period from the current sub-period to the last sub-period in the target period.

[0168] Continuing with the example of a target time period comprising multiple sub-time periods A, B, C, and D in sequential order, the energy scheduling model can predict the scheduling strategy corresponding to the target time period based on the predicted working parameters and target data at time t0. This scheduling strategy includes sub-strategies corresponding to each sub-time period, such as A, B, C, and D. The energy scheduling model can then predict the sub-strategies corresponding to B, C, D, and the next sub-time period of D based on the predicted working parameters and target data at time t1. After predicting the sub-strategy corresponding to time period B based on the predicted working parameters and target data at time t1, the scheduling strategy can be updated based on this newly obtained sub-strategy, thus obtaining the final scheduling strategy.

[0169] Step 130: Input the sub-strategies corresponding to the sub-time periods into the first simulation model in sequence to obtain the sub-execution results corresponding to each sub-time period;

[0170] In this step, the first simulation model is a pre-constructed electrical simulation model, mathematical model, or a comprehensive model combining the electrical simulation model and the mathematical model, used to simulate the overall target application object and the operating status of each device in the target application object.

[0171] It is understandable that the first simulation model may differ for different target applications, and it can be constructed according to the actual situation during the actual execution process.

[0172] Figure 5 An example of a first simulation model is provided, which includes photovoltaic equipment, energy storage equipment, charging piles, and other equipment.

[0173] Continue to refer to Figure 2 After obtaining the sub-strategy corresponding to the sub-time period, the sub-strategy can be sent to the first simulation model, which will then execute the sub-strategy and obtain the corresponding sub-execution result.

[0174] The sub-execution results may include, but are not limited to, information such as the working status of the relevant devices in the target application object after executing the sub-strategy corresponding to the sub-period, the electricity cost generated, the carbon emissions, and the benefits.

[0175] In actual execution, the first simulation model executes the sub-strategies of each sub-time period sequentially according to the order of the sub-time periods until all sub-strategies corresponding to the sub-time periods within the target time period are executed, thus ending one round of simulation data loop. Through one round of simulation data loop, the sub-execution result corresponding to each sub-time period can be obtained. Combining the above sub-execution results, the final simulation result of the first simulation model after executing all sub-strategies corresponding to the sub-time periods included in the target time period can be obtained in the current simulation data loop.

[0176] The simulation results can characterize the total electricity cost, total carbon emissions, and total revenue generated by the first simulation model after executing all the sub-strategies corresponding to the sub-time periods included in the target time period.

[0177] Understandably, the core of the first simulation model is a mathematical solver, which has no concept of time. Therefore, it can quickly execute all sub-policies within the target time period in a short time, thereby realizing real-time deduction of the scheduling policy and obtaining simulation results.

[0178] Step 140: Based on the simulation results obtained from the sub-execution results corresponding to each sub-period in the target time period, adjust the model parameters of the energy scheduling model, and re-execute the steps of obtaining the prediction working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period until the preset stopping condition is reached.

[0179] In this step, the simulation results can characterize the total electricity cost, total carbon emissions, and total revenue generated by the first simulation model after executing the sub-strategies corresponding to all sub-time periods included in the target time period.

[0180] After completing one small loop of simulation data, the next small loop of simulation data can be entered, which is the execution of the large loop of simulation data; continue to refer to Figure 2 After obtaining the simulation results corresponding to the target time period, the energy scheduling model can be fine-tuned by adjusting the model parameters. With the input features remaining unchanged, the sub-strategies output by the energy scheduling model will also be different as the model parameters change. The simulation results obtained by the first simulation model executing each sub-strategy may also be different.

[0181] The preset stopping condition can be the maximum number of times steps 110 to 140 are repeated, which is the maximum number of loops of the simulation data loop. This can be user-defined or set according to the performance of the first simulation model. As long as it ensures that all sub-strategies included in the target time period can be executed within the sub-time period given the number of servers available to the user.

[0182] The preset stopping condition can also be set as the maximum number of times the model parameters can be adjusted. The maximum number of adjustments can be set based on the performance of the simulation model. As long as the number of servers available to the user is sufficient, all sub-strategies can be executed and simulation results can be obtained within the target sub-period.

[0183] During the adjustment process, it is essential to ensure that the model parameters after each adjustment are different from the previous model parameters, thereby obtaining different energy scheduling algorithms. Using the adjusted energy scheduling model, steps 110 to 140 are repeated to perform a large-scale simulation data loop, in order to obtain multiple different scheduling strategies and the simulation results of the first simulation model after executing each scheduling strategy, until the preset stopping condition is reached, at which point the large-scale simulation data loop ends.

[0184] It is understood that steps 110 to 140 above can be performed at any time of day, such as before the day (quasi-real-time) or during the day (real-time), and this application does not limit this.

[0185] In actual implementation, a B / S architecture can be adopted, based on the Java language and Spring Cloud framework. Alternatively, a C / S architecture or other architectures, as well as other languages ​​(such as C++) and technical frameworks, can be used as needed. These will not be elaborated on here.

[0186] Step 150: Based on the simulation results, select the target scheduling strategy corresponding to the target application object.

[0187] In this step, the target scheduling strategy can be an energy scheduling strategy selected from multiple scheduling strategies obtained from multiple rounds of simulation data loops, which is provided for reference or execution by the target application object within the target time period.

[0188] In some embodiments, the target scheduling strategy can be the scheduling strategy that maximizes total revenue, or the scheduling strategy that minimizes carbon emissions, or the scheduling strategy that minimizes costs, or the best scheduling strategy after comprehensive evaluation of multiple indicators, etc.

[0189] In some embodiments, the target scheduling policy can also be determined by user input, such as by the user selecting one or more scheduling policies from multiple scheduling policies as the target scheduling policy according to actual needs.

[0190] like Figure 3 As shown, in some embodiments, after step 150, the method may further include: distributing the target scheduling strategy to the energy scheduling system so that the target application object can perform actual energy scheduling according to the target scheduling strategy during the target time period.

[0191] It is understandable that the target scheduling strategy may also be updated after the target time period is updated. For example, after determining the target scheduling strategy between 0:00 and 24:00 on the same day based on the relevant data at the current time, after the end of the day, the relevant data for the next day can be obtained again and steps 110 to 150 can be repeated to obtain the target scheduling strategy for the next day, so as to schedule and control the work of the target application object on the next day.

[0192] In this application, an energy dispatch model is used to predict sub-strategies for each sub-period based on the predicted operating parameters and target data. A pre-built first simulation model corresponding to the target application object executes each sub-strategy, simulating the entire target period to obtain simulation results. By adjusting the model parameters of the energy dispatch model and performing multiple simulations, multiple different dispatch strategies corresponding to the target period and different simulation results obtained after the first simulation model executes each dispatch strategy can be obtained. Based on the simulation results, a target dispatch strategy is selected from multiple different dispatch strategies for the target application object to use as a dispatch reference or for execution. This comprehensively considers factors such as the real-time changes in the power generation capacity, electricity load, grid supply capacity and price, and energy storage capacity of the target application object. It adapts to the requirement of "real-time or near-real-time updates" brought about by various complex factors such as seasonal changes, weather changes, and changes in user energy consumption scenarios, continuously providing users with dispatch strategies that meet their needs as much as possible, thereby improving the user experience.

[0193] According to the energy scheduling method provided in the embodiments of this application, the energy scheduling model generates sub-strategies based on predicted short-term data, and the first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target scheduling strategy that better meets user needs. This method is applicable to complex and ever-changing scheduling scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with scheduling strategies that meet their needs as much as possible, thereby improving the user experience.

[0194] In some embodiments, step 110 may include:

[0195] Based on the forecasting parameters corresponding to the previous sub-period and the power forecasting data in the target data, the forecasting parameters corresponding to the current sub-period are obtained.

[0196] In this embodiment, the prediction parameters corresponding to the current sub-period can be predicted by a simulation system or a pre-trained network model based on the prediction parameters corresponding to the previous sub-period and the power prediction data in the target data.

[0197] The energy scheduling method provided in the embodiments of this application predicts the working parameters of the next sub-period by predicting the working parameters of the previous sub-period. It has high real-time performance and can solve the problem of not being able to obtain the real-time working parameters of the time that has not yet occurred. Thus, it simulates the real scheduling environment and is suitable for complex and ever-changing scheduling environments, improving the accuracy and real-time performance of the obtained scheduling strategy.

[0198] Continue to refer to Figure 2 In some embodiments, the prediction parameters for the current sub-period are obtained based on the prediction parameters corresponding to the previous sub-period and the power prediction data in the target data. This may include:

[0199] The prediction parameters corresponding to the previous sub-period and the power prediction data in the target data are input into the second simulation model to obtain the prediction parameters corresponding to the current sub-period.

[0200] In this embodiment, the second simulation model is a pre-constructed electrical simulation model, mathematical model, or a comprehensive model combining an electrical simulation model and a mathematical model, used to simulate the overall target application object and the operating status of each device in the target application object.

[0201] The device models included in the first simulation model and the second simulation model may overlap. The first simulation model is used to execute sub-strategies, and the second simulation model is used to predict real-time operating parameters.

[0202] In practical applications, for the first sub-period in the target time period, if the actual working parameters of the first sub-period can be obtained, the actual working parameters can be determined as the predicted working parameters corresponding to the first sub-period. The actual working parameters corresponding to the first sub-period and the power prediction data are input into the second simulation model, and the predicted working parameters corresponding to the next sub-period of the first sub-period are obtained by simulation by the second simulation model. The predicted working parameters can be used as the real-time working parameters corresponding to the next sub-period.

[0203] For example, based on the actual working parameters and power forecast data corresponding to time period A, the predicted working parameters corresponding to time period B are obtained through simulation. Then, based on the predicted working parameters and power forecast data corresponding to time period B, the predicted working parameters corresponding to time period C are obtained through simulation, and so on, until the predicted working parameters corresponding to the last sub-time period of the target time period are predicted.

[0204] According to the energy scheduling method provided in the embodiments of this application, the real-time data of the next sub-period is obtained by simulating the real-time data of the previous sub-period. The scheduling strategy is predicted based on the real-time data of each sub-period obtained by simulation. It has high real-time performance, can simulate the real scheduling environment, is suitable for complex and ever-changing scheduling environments, and improves the accuracy and real-time performance of the obtained scheduling strategy.

[0205] In some embodiments, obtaining the prediction parameters for the current sub-period based on the prediction parameters corresponding to the previous sub-period and the power prediction data in the target data may include:

[0206] The second simulation model is adjusted based on the sub-execution result corresponding to the previous sub-period of the current sub-period.

[0207] The prediction parameters corresponding to the previous sub-period and the power prediction data in the target data are input into the adjusted second simulation model to obtain the prediction parameters corresponding to the current sub-period.

[0208] In this embodiment, the sub-execution result corresponding to the previous sub-period is obtained by the first simulation model through simulation deduction based on the sub-strategy corresponding to the previous sub-period; the sub-strategy corresponding to the previous sub-period is predicted by the energy scheduling model through the prediction working parameters and target data corresponding to the previous sub-period.

[0209] The second simulation model is dynamically updated based on the sub-execution results corresponding to the previous sub-period.

[0210] The following example illustrates the concept of a target time period comprising multiple sub-time periods, such as A, B, C, and D, arranged in a sequential order, with time period C being the current sub-time period and time period B being the previous sub-time period.

[0211] In actual implementation, the prediction working parameters and power prediction data corresponding to time period A are input into the second simulation model for prediction, and the prediction working parameters corresponding to time period B can be obtained.

[0212] By inputting the prediction parameters and target data corresponding to time period B into the energy scheduling model to predict the scheduling strategy, the sub-strategy corresponding to time period B can be obtained.

[0213] By inputting the sub-strategy corresponding to time period B into the first simulation model for execution, the sub-execution result corresponding to time period B can be obtained.

[0214] The second simulation model is adjusted by the sub-execution results corresponding to time period B. For example, the working status and working parameters of each device in the second simulation model are adjusted to the working status and working parameters after the execution of the sub-strategy corresponding to time period B. For example, the device model corresponding to the energy storage device changes from charging to discharging state, and the power grid draw is reduced.

[0215] The second simulation model is driven to continue generating the prediction parameters for the next sub-period. For example, if the prediction parameters and power prediction data for period B are input into the adjusted second simulation model for prediction, the prediction parameters for period C can be obtained.

[0216] According to the energy scheduling method provided in the embodiments of this application, the second simulation model of the current sub-period is dynamically updated by the sub-execution result corresponding to the previous sub-period. This can more realistically simulate the dynamic changes in the working status of each device during the actual scheduling process of the target application object, improve the accuracy and realism of the simulation results, and better adapt to the requirement of "real-time or near-real-time updates" brought about by various complex factors such as seasonal changes, weather changes, and changes in user energy consumption scenarios.

[0217] In some embodiments, step 120 may include:

[0218] The second simulation model is used to simulate and deduce the prediction working parameters corresponding to the previous sub-period of each sub-period until the prediction working parameters corresponding to each sub-period included in the target period are obtained.

[0219] The energy scheduling model processes the predicted working parameters and target data for each sub-period within the target time period to obtain the scheduling strategy corresponding to the target time period. The scheduling strategy includes the sub-strategies corresponding to each sub-period.

[0220] In this embodiment, the target time period includes multiple sub-time periods such as A, B, C, and D in sequential order, with time period C being the current sub-time period and time period B being the previous sub-time period.

[0221] Continue to refer to Figure 2 In actual execution, the prediction working parameters and power prediction data corresponding to time period A are input into the second simulation model for prediction, and the prediction working parameters corresponding to time period B can be obtained.

[0222] By inputting the prediction parameters and target data corresponding to time period B into the energy scheduling model to predict the scheduling strategy, the sub-strategy corresponding to time period B can be obtained.

[0223] By inputting the prediction working parameters and power prediction data corresponding to time period B into the second simulation model for prediction, the prediction working parameters corresponding to time period C can be obtained.

[0224] By inputting the prediction parameters and target data corresponding to time period C into the energy scheduling model to predict the scheduling strategy, the sub-strategy corresponding to time period C can be obtained; and so on, until the sub-strategy corresponding to all time periods is obtained; by sorting the sub-strategies in chronological order, the scheduling strategy corresponding to the target time period can be obtained.

[0225] According to the energy scheduling method provided in the embodiments of this application, the predicted working parameters of the current sub-period are predicted from the predicted working parameters of the previous sub-period through the second simulation model, and are used as the real-time working parameters of the current sub-period. The energy scheduling model predicts the sub-strategy of the current sub-period based on the real-time working parameters of the current sub-period. It can quickly and efficiently simulate and deduce the working status of different sub-periods within the target period, and has high real-time performance. It can adapt to the requirements of "real-time or near-real-time updates" brought about by various complex factors such as seasonal changes, weather changes, and changes in user energy consumption scenarios.

[0226] In some embodiments, step 120 may include:

[0227] For each sub-period, that sub-period is taken as the current sub-period, and the first simulation model is adjusted according to the sub-execution result corresponding to the current sub-period;

[0228] The sub-strategy corresponding to the next sub-period of the current sub-period is input into the adjusted first simulation model to obtain the sub-execution result corresponding to the next sub-period of the current sub-period.

[0229] In this embodiment, the first simulation model is dynamically updated based on the sub-execution result corresponding to the previous sub-period.

[0230] Let's continue with the example of the target time period, which includes multiple sub-time periods such as A, B, C, and D in sequential order.

[0231] In actual implementation, the prediction working parameters and power prediction data corresponding to time period A are input into the second simulation model for prediction, and the prediction working parameters corresponding to time period B can be obtained.

[0232] By inputting the prediction parameters and target data corresponding to time period B into the energy scheduling model to predict the scheduling strategy, the sub-strategy corresponding to time period B can be obtained.

[0233] By inputting the sub-strategy corresponding to time period B into the first simulation model for execution, the sub-execution result corresponding to time period B can be obtained.

[0234] Adjust the first and second simulation models by the sub-execution results corresponding to time period B. For example, adjust the working status and working parameters of each device in the first and second simulation models to the working status and working parameters after the execution of the sub-strategy corresponding to time period B.

[0235] By inputting the prediction working parameters and power prediction data corresponding to time period B into the adjusted second simulation model for prediction, the prediction working parameters corresponding to time period C can be obtained.

[0236] By inputting the prediction parameters and target data corresponding to time period C into the energy scheduling model to predict the scheduling strategy, the sub-strategy corresponding to time period C can be obtained.

[0237] By inputting the sub-strategy corresponding to time period C into the adjusted first simulation model for execution, the sub-execution result corresponding to time period C can be obtained.

[0238] Then, based on the sub-execution results corresponding to time period C, the model parameters of the first and second simulation models are updated, and so on, until the sub-strategy of the last sub-time period is executed.

[0239] According to the energy scheduling method provided in the embodiments of this application, the first simulation model of the current sub-period is dynamically updated by the sub-execution result corresponding to the previous sub-period. This can more realistically simulate the dynamic changes in the working status of each device during the actual scheduling process of the target application object, improve the accuracy and realism of the simulation results, and better adapt to the requirements of "real-time or near-real-time updates" brought about by various complex factors such as seasonal changes, weather changes, and changes in user energy consumption scenarios on the scheduling strategy.

[0240] In some embodiments, step 120 may include:

[0241] The sub-strategy corresponding to the sub-time period is input into the first simulation model so that at least one target device model in the first simulation model can execute the sub-strategy and obtain the sub-execution result corresponding to the sub-time period.

[0242] In this embodiment, the target device can be a simulation model corresponding to any device in the target application, such as a device model corresponding to an energy storage device, a device model corresponding to a power generation device, and a device model corresponding to a power consumption device. During actual execution, the target device model can be determined based on the execution objects involved in the sub-strategy. For example, if the sub-strategy needs to adjust the charging and discharging state of the energy storage device and adjust the power generation capacity of the power generation device, then the target device model can be determined as the device model corresponding to both the energy storage device and the power generation device.

[0243] The sub-execution result includes the target device model and the simulation working parameters of the device models associated with the target device model after the sub-execution result is executed. It can be understood that for a simulation system with a topology structure, when the target device executes a sub-strategy and causes a change in its working state, the working states of some devices associated with the target device may change synchronously, thus obtaining new simulation working parameters.

[0244] The simulation parameters are used to adjust the first simulation model so that the adjusted first simulation model can be used to simulate and deduce the sub-execution result corresponding to the next sub-period. For example, if the device model corresponding to the energy storage device changes from charging to discharging, then in the next round of simulation, the device model corresponding to the energy storage device will use the discharging state as the initial state to execute the sub-strategy corresponding to the next sub-period.

[0245] According to the energy scheduling method provided in the embodiments of this application, by constructing a simulation model to execute the sub-strategies corresponding to each sub-time period in chronological order, the simulation model can be dynamically updated based on the sub-execution results of the previous sub-time period, driving the updated simulation model to execute the sub-strategy corresponding to the current sub-time period. It has a high adaptive adjustment capability, thereby more realistically simulating the linkage between various devices and the complex and ever-changing environmental conditions of the target application object in actual operation, improving the accuracy and real-time performance of the simulation results, and meeting the needs of users.

[0246] In some embodiments, step 140 may include:

[0247] Perform the following operations sequentially according to the order of the sub-time periods:

[0248] S1. Initialize the current sub-time period;

[0249] S2. Obtain the prediction parameters and target data corresponding to the current sub-period;

[0250] S3. Input the prediction working parameters and target data corresponding to the current sub-period into the energy scheduling model to obtain the sub-strategy corresponding to the current sub-period.

[0251] S4. Input the sub-strategy corresponding to the current sub-period into the first simulation model to obtain the sub-execution result corresponding to the sub-strategy corresponding to the current sub-period.

[0252] S5. If the current sub-period is not the last sub-period of the target period, then return to execute S1;

[0253] S6. If the current sub-period is the last sub-period of the target period, then terminate the operation.

[0254] In this embodiment, the example will continue to be used, which includes multiple sub-time periods such as A, B, C and D in sequential order.

[0255] Continue to refer to Figure 2 By inputting the prediction working parameters and power prediction data corresponding to time period A into the second simulation model for prediction, the prediction working parameters corresponding to time period B can be obtained.

[0256] By inputting the prediction parameters and target data corresponding to time period B into the energy scheduling model to predict the scheduling strategy, the sub-strategy corresponding to time period B can be obtained.

[0257] By inputting the sub-strategy corresponding to time period B into the first simulation model for execution, the sub-execution result corresponding to time period B can be obtained.

[0258] Adjust the first and second simulation models by the sub-execution results corresponding to time period B. For example, adjust the working status and working parameters of each device in the first and second simulation models to the working status and working parameters after the execution of the sub-strategy corresponding to time period B.

[0259] If time period B is not the last sub-time period, it is assumed that all sub-strategies within the target time period have not been executed. Therefore, the second simulation model is driven to continue generating the prediction working parameters for the next sub-time period. The prediction working parameters corresponding to time period B and the power prediction data are input into the adjusted second simulation model for prediction, and the prediction working parameters corresponding to time period C can be obtained. Time period C is initialized as the current sub-time period.

[0260] By inputting the prediction parameters and target data corresponding to time period C into the energy scheduling model to predict the scheduling strategy, the sub-strategy corresponding to time period C can be obtained.

[0261] By inputting the sub-strategy corresponding to time period C into the adjusted first simulation model for execution, the sub-execution result corresponding to time period C can be obtained.

[0262] Then, the model parameters of the first and second simulation models are updated based on the sub-execution results corresponding to time period C;

[0263] Since period C is not the last sub-period, it is assumed that all sub-strategies within the target period have not been executed. Therefore, the second simulation model is driven to continue generating the prediction working parameters for the next sub-period. The prediction working parameters and power prediction data corresponding to period C are then input into the adjusted second simulation model for prediction, and the prediction working parameters corresponding to period D can be obtained. Period D is then initialized as the current sub-period.

[0264] By inputting the prediction parameters and target data corresponding to time period D into the energy scheduling model to predict the scheduling strategy, the sub-strategy corresponding to time period D can be obtained.

[0265] By inputting the sub-strategy corresponding to time period D into the adjusted first simulation model for execution, the sub-execution result corresponding to time period D can be obtained.

[0266] If time period D is the last sub-time period, then it is considered that all sub-strategies within the target time period have been executed, and the operation is terminated. The simulation results corresponding to the target time period are obtained based on the sub-execution results of each sub-time period, the current loop ends, and the next large loop of simulation data begins.

[0267] According to the energy scheduling method provided in the embodiments of this application, by constructing two loop frameworks, a small loop and a large loop of simulation data, it is possible to obtain simulation results that are closer to the current user needs, and then select target scheduling strategies for reference or execution by the target application objects. This can meet the needs of real-time or near-real-time operation and make the scheduling results more in line with the current needs of the user.

[0268] In some embodiments, step 140 may include:

[0269] If the number of adjustments is less than or equal to the first threshold, the steps of obtaining the current sub-period and the prediction working parameters and target data corresponding to each remaining sub-period in the target period are re-executed.

[0270] If the number of adjustments exceeds the first threshold, the steps of obtaining the current sub-period and the prediction working parameters and target data corresponding to each remaining sub-period in the target time period are stopped.

[0271] In this embodiment, the number of adjustments refers to the number of times the model parameters are adjusted, with each large cycle of simulation data corresponding to one adjustment of the model parameters. The first threshold can be user-defined or set based on the performance of the simulation model, such as to ensure that, given the number of servers available to the user, all sub-strategies within the target time period can be executed within a sub-time period.

[0272] Let's continue with the example of the target time period, which includes multiple sub-time periods such as A, B, C, and D in sequential order.

[0273] like Figure 2As shown, in the actual execution process, the prediction parameters and power prediction data corresponding to time period A are input into the second simulation model for prediction, resulting in the prediction parameters for time period B. The prediction parameters and target data corresponding to time period B are input into the energy dispatch model for dispatch strategy prediction, resulting in the sub-strategy for time period B. The sub-strategy for time period B is input into the first simulation model for execution, resulting in the sub-execution result for time period B. The first and second simulation models are adjusted based on the sub-execution result for time period B. The prediction parameters and power prediction data for time period B are then input into the adjusted second simulation model for prediction, yielding the desired result. The prediction parameters for time period C are obtained; the prediction parameters and target data for time period C are input into the energy scheduling model to predict the scheduling strategy, and the sub-strategy for time period C is obtained; the sub-strategy for time period C is input into the adjusted first simulation model for execution, and the sub-execution result for time period C is obtained; the model parameters of the first and second simulation models are updated according to the sub-execution result for time period C, and so on, until the sub-strategy for time period D is executed. It is considered that all sub-strategies within the target time period have been executed. Then, according to the sub-execution result for each sub-time period, the simulation result X1 and the scheduling strategy Y1 for the target time period are obtained.

[0274] Then, the simulation data loop continues, the current sub-period is initialized, time period A is reset as the current sub-period, and the model parameters of the energy scheduling model are adjusted to obtain a new energy scheduling model. The number of adjustments is incremented by one.

[0275] If the number of executions in the simulation data loop exceeds the upper limit, and the number of adjustments is less than or equal to the first threshold, it is considered not to exceed the upper limit. Then, the prediction parameters and power prediction data for time period A are input into the second simulation model for prediction, yielding the prediction parameters for time period B. The prediction parameters and target data for time period B are then input into a new energy dispatch model for dispatch strategy prediction, yielding the sub-strategy for time period B. The sub-strategy for time period B is then input into the first simulation model for execution, yielding the sub-execution result for time period B. This process continues until the sub-strategy for time period D is executed. Based on the sub-execution results for each sub-time period in this loop, the simulation result X2 for the target time period and the dispatch strategy Y2 are obtained.

[0276] Then, the simulation data loop continues, the current sub-period is initialized, time period A is reset as the current sub-period, and the model parameters of the energy scheduling model are adjusted to obtain a new energy scheduling model. The number of adjustments is incremented by one.

[0277] Determine whether the number of times the simulation data loop is executed exceeds the upper limit. If the number of times is adjusted is still less than or equal to the first threshold, it is considered that the upper limit has not been exceeded. Then repeat the above steps to obtain the simulation result X3 and scheduling strategy Y3 corresponding to the target time period.

[0278] If the number of adjustments exceeds the first threshold, it is considered to have exceeded the upper limit, and the simulation data loop ends.

[0279] After the simulation data loop ends, based on the simulation results X1, X2, X3, etc. obtained from each loop, the simulation result that is closer to the current user needs is selected, and the scheduling strategy corresponding to the simulation result that is closer to the current user needs is determined as the target scheduling strategy.

[0280] According to the energy scheduling method provided in the embodiments of this application, by constructing two loop frameworks, a small loop and a large loop of simulation data, it is possible to obtain simulation results that are closer to the current user needs, and then select target scheduling strategies for reference or execution by the target application objects. This can meet the needs of real-time or near-real-time operation and make the scheduling results more in line with the current needs of the user.

[0281] In some embodiments, adjusting the model parameters of the energy dispatch model may include at least one of the following methods:

[0282] The model parameters of the energy dispatch model are adjusted with the goal of obtaining the optimal value of the simulation results;

[0283] or,

[0284] A subset of parameters from the energy dispatch model were randomly selected and determined as the parameters to be adjusted.

[0285] Adjust the parameter to be adjusted so that the adjusted parameter is different from the parameter after each previous adjustment;

[0286] or,

[0287] The model parameters of the energy scheduling model are adjusted using a tuner.

[0288] or,

[0289] The model parameters of the energy dispatch model are adjusted using a loss function; the loss function is constructed based on simulation results and target simulation results.

[0290] In this embodiment, an AI algorithm-based tuner can be used to randomly adjust the model parameters of the energy scheduling model to obtain a new energy scheduling model.

[0291] For example, optimization algorithms, such as loss functions, can be used to adjust the model parameters of the energy scheduling model.

[0292] In some embodiments, during each adjustment process, some parameters of the energy dispatch model may be randomly selected as parameters to be adjusted, and the selected parameters to be adjusted may be adjusted.

[0293] It should be noted that during each adjustment process, the adjusted model parameters should not be completely the same as the previous model parameters. That is, each new energy scheduling model (or energy scheduling algorithm) should not be completely the same as the previous energy scheduling model (or energy scheduling algorithm), so as to be able to predict different sub-policies.

[0294] In actual implementation, the model parameters can be fine-tuned in any feasible way so that each energy scheduling model can output different scheduling strategies. The first simulation model executes different scheduling strategies to generate different simulation results, thereby selecting the optimal simulation result from multiple simulation results and determining the scheduling strategy corresponding to the optimal simulation result as the target scheduling strategy.

[0295] The energy scheduling method provided in the embodiments of this application can obtain a variety of different energy scheduling algorithms by fine-tuning the model parameters, and can quickly obtain a variety of different scheduling strategies for selection. The scheduling strategies are simulated and deduced by combining the simulation model, thereby obtaining a scheduling strategy that is as close as possible to the user's needs. It has high flexibility and is simple to operate and easy to implement.

[0296] In some embodiments, step 150 may include:

[0297] Based on the simulation results, calculate the evaluation score corresponding to the simulation results from at least one dimension;

[0298] The scheduling strategy corresponding to the simulation result with the highest evaluation score is determined as the target scheduling strategy.

[0299] In this embodiment, the evaluation score is used to characterize the degree of closeness between the simulation result and the user's goal. The higher the evaluation score, the closer the simulation result corresponding to the evaluation score is to the user's needs and the scheduling strategy corresponding to the simulation result. The dimensions may include, but are not limited to, benefits, costs, energy consumption, carbon emissions, and ease of implementation.

[0300] In some embodiments, revenue may include revenue corresponding to each sub-period or total revenue corresponding to the target period.

[0301] In some embodiments, at least one dimension may include the total revenue corresponding to the target time period. Optimizing to maximize total revenue can effectively guarantee user operational profitability.

[0302] Continuing with the example of a target time period comprising multiple sub-time periods, A, B, C, and D in sequential order; for example... Figure 2 As shown, for each large loop of simulation data, a set of scheduling strategies corresponding to the target time period and the total benefit of the simulation system under the execution of the scheduling strategy can be obtained. The maximum benefit is selected from these benefits, and the scheduling strategy corresponding to the maximum benefit is determined as the target scheduling strategy.

[0303] In some embodiments, continue to refer to Figure 2 Furthermore, the target scheduling strategy can be defined as the actual scheduling strategy for the target application on that day and sent to the energy scheduling system for execution, so as to perform energy scheduling for the target application.

[0304] According to the energy dispatching method provided in the embodiments of this application, the dispatching strategy is simulated to obtain the simulation results corresponding to the dispatching strategy. Based on the simulation inference capability, the simulation results are evaluated according to user needs and one or more dimensions. With the goal of selecting the one that is closest to the current needs of the user, various prediction algorithms and dispatching algorithms in the energy dispatching model are trained and optimized. The dispatching strategy and model parameters that are closest to the current needs of the user under the influence of multiple factors such as power generation capacity, power load, grid power supply capacity and electricity price in the target time period are obtained. This enables real-time or near real-time high-frequency iteration, self-iteration and self-optimization, adapting to complex and ever-changing dispatching environments and meeting user needs.

[0305] like Figure 3 As shown, in some embodiments, after step 150, the method may further include:

[0306] The model parameters corresponding to the target scheduling strategy are determined as the actual model parameters of the energy scheduling model during the target time period.

[0307] In this embodiment, when the target scheduling strategy is determined to be the strategy actually executed by the target application object, the model parameters of the predicted target scheduling strategy can be sent to the energy scheduling system so that the model parameters corresponding to the target scheduling strategy can be used as the actual model parameters of the energy scheduling model in the target time period for the actual execution of the scheduling algorithm.

[0308] For example, if the target time period is from 0:00 to 24:00 on the same day, the target scheduling strategy can be sent to the energy scheduling model as a model parameter in the actual working process of the energy scheduling model on the same day.

[0309] According to the energy scheduling method provided in the embodiments of this application, by determining the model parameters corresponding to the target scheduling strategy as the actual model parameters of the energy scheduling model in the target time period, the energy scheduling model can be updated in the short term with the goal of optimizing the simulation results. This makes the prediction results of the energy scheduling model in the short term more in line with the user's short-term needs, and the model parameters of the energy scheduling model can be dynamically adjusted according to changes in user needs, thus having high flexibility.

[0310] Continue to refer to Figure 3 In some embodiments, after step 150, the method may further include:

[0311] Optimize model parameters based on the target time interval.

[0312] In this embodiment, the target time interval is longer than the target period, such as one month, three months, or half a year.

[0313] The algorithm inference engine, algorithm OTA, and scheduling algorithm execution feedback of the energy scheduling model can be trained and optimized in any feasible way.

[0314] During the optimization process, the data fed back to the "training dataset" by the "scheduling algorithm execution feedback" module may include the following:

[0315] 1) Daily two-part electricity pricing data, including peak-valley pricing, capacity pricing, and demand pricing;

[0316] 2) Incoming line data for different time periods on the same day;

[0317] 3) Load data for different time periods on the same day;

[0318] 4) Power generation data for different time periods on the same day;

[0319] 5) Energy storage SOC and charge / discharge power data for different time periods on the same day;

[0320] 6) Scheduling strategies for different time periods on the same day;

[0321] 7) Power factor for the day;

[0322] 8) Total energy dispatch revenue for the day, as well as individual revenue items, including capacity charge revenue, demand charge revenue, and revenue from circumventing penalties.

[0323] Based on the high-frequency optimization of the energy scheduling model using the simulation system, and combined with the feedback from the actual execution of each target scheduling strategy by the energy scheduling system, the energy scheduling model is optimized at a low frequency using an optimization algorithm, thereby further improving the accuracy of the energy scheduling model and the prediction results.

[0324] like Figure 6 As shown, in actual execution, the energy dispatch system can send the target dispatch strategy generated by the energy dispatch model yesterday, including the optimal daytime dispatch strategy yesterday, all daytime dispatch strategies actually executed yesterday, and minute-level current, voltage, and power data of all source-grid-load-storage devices of the microgrid yesterday, to the algorithm training engine of the energy dispatch algorithm at dawn every day.

[0325] The algorithm training engine incorporates the received day-ahead and mid-day scheduling strategies and the minute-level current, voltage, and power data actually generated by the microgrid devices into the training and validation sets of the energy scheduling algorithm.

[0326] After the target time interval, the algorithm training engine trains the energy scheduling model based on the existing training and validation sets to generate a new energy scheduling model.

[0327] The new energy scheduling model is released to the energy scheduling system, and the low-frequency optimization of the energy scheduling algorithm is completed.

[0328] The energy scheduling method provided in this application embodiment improves the accuracy of the energy scheduling model and enhances the prediction results by performing simulations based on the changes in the target application object itself, thereby improving the scheduling effect.

[0329] In some embodiments, prior to step 110, the method may further include:

[0330] Based on at least one of the following: equipment information of the target application object, local control strategies corresponding to the equipment included in the target application object, power topology information of the target application object, and overall coordination control strategy of the target application object, a first simulation model and a second simulation model are constructed.

[0331] In this embodiment, the device information is used to characterize the device categories, number of devices, and attributes and basic parameters of each device included in the target application object, including but not limited to: energy storage devices, power generation devices, lighting devices, air conditioning devices, and other devices.

[0332] Local control strategies are strategies for controlling the devices to be controlled, such as black-start control algorithms, photovoltaic tracking strategies, and energy storage charging and discharging control strategies.

[0333] Power topology is used to characterize the connection relationships between various devices. Different power plants correspond to different power topologies. In the actual construction process, simulation models should be built one by one based on each power plant according to different application scenarios.

[0334] In some embodiments, Simulink or other simulation modeling tools can be used to model the above content to obtain a first simulation model and a second simulation model.

[0335] When modeling a microgrid / energy station, the following should be included:

[0336] 1) Perform electrical modeling for each device that makes up the microgrid / energy station;

[0337] 2) Model and develop local control strategies for each device in the microgrid / energy station;

[0338] 3) Model the power topology of the microgrid / energy station;

[0339] 4) Model and develop the overall coordinated control strategy for microgrids / energy stations.

[0340] After the simulation modeling is completed, the built simulation model can be compiled using Simulink or other simulation modeling tools to generate an executable file (such as an .exe file under Windows or an .o file under Linux). This executable file can be used as a "simulation executable program" output by the simulation design tool and output to the next stage.

[0341] According to the energy dispatching method provided in this application embodiment, the target application object is simulated and modeled based on its actual characteristics. This yields a simulation model that simulates the local control strategies of each device within the target application object and the overall coordinated control strategy of the target application object. This provides real-time and near-real-time simulation and deduction capabilities, enabling the energy dispatching algorithm to track changes in the target application object's power generation, power consumption, energy storage, and grid power supply status in real-time and near-real-time, as well as changes in information such as photovoltaic grid-connected electricity prices and electricity prices. This improves the optimization effect of the energy dispatching algorithm, allowing it to generate dispatching strategies that are closer to the user's current needs based on the real-time changes of the target application object, thus meeting the user's requirements.

[0342] In some embodiments, constructing a first simulation model and a second simulation model based on at least one of the following: device information of the target application object, local control strategies corresponding to the devices included in the target application object, power topology information of the target application object, and overall coordinated control strategies of the target application object, may include:

[0343] Based on the equipment information of the target application object, the local control strategy corresponding to the equipment included in the target application object, the power topology information of the target application object, and the overall coordination control strategy of the target application object, construct the equipment model corresponding to the site equipment and the site model corresponding to the energy site.

[0344] Based on at least one of the equipment model and the site model, a first simulation model and a second simulation model are obtained.

[0345] In this embodiment, the equipment model corresponding to the site equipment is the simulation model corresponding to each equipment included in the target application object, including electrical models and / or digital models, etc., wherein the equipment includes, but is not limited to: each energy storage device, each power generation device, each power consumption device, and the power grid, etc.

[0346] The site model corresponding to the energy site is a complete simulation model that includes at least some of the equipment, as well as the topology and relationships between at least some of the equipment, including electrical models and / or digital models, etc.

[0347] In some embodiments, based on the device information of the target application object, the local control strategy corresponding to the devices included in the target application object, the power topology information of the target application object, and the overall coordinated control strategy of the target application object, a device model corresponding to the site devices and a site model corresponding to the energy site are constructed, which may include:

[0348] Based on the device information of the target application object and the local control strategy corresponding to the devices included in the target application object, construct the device model corresponding to the site device;

[0349] Based on the power topology information of the target application object and the overall coordination and control strategy of the target application object, a site model corresponding to the energy site is constructed.

[0350] After constructing the equipment model and the site model, one or more of them can be selected as the first simulation model and the second simulation model.

[0351] In some embodiments, the equipment model and the site model can be combined to obtain a first simulation model and a second simulation model.

[0352] like Figure 7 As shown, in actual execution, electrical models or equivalent models of power generation equipment, energy storage equipment, and power consumption equipment can be defined, modeled by simulation design tools, and the simulation models are compiled into online simulation executable programs to obtain the first simulation model and the second simulation model, which facilitates online simulation in subsequent applications.

[0353] In some embodiments, containerization technology can be used to encapsulate the "emulated executable program" into a container, and then schedule its execution through a container scheduling system, such as using Docker+K8S containerization technology or other mainstream containerization technologies in the industry, to ensure the concurrent execution and high reliability of the "emulated executable program".

[0354] When the "simulation executable program" is executed in real time, it can generate real-time simulation data for each device in the target application object, including but not limited to: real-time voltage and real-time current of each device, or it can output other data as needed, such as power and electricity consumption.

[0355] The data output by the "simulation executable program" can be uploaded in real time to the energy scheduling strategy deduction subsystem via Modbus, MQTT or other protocols, driving the real-time or near-real-time self-optimization and self-iteration of energy scheduling algorithms and strategies in the energy scheduling model.

[0356] like Figure 4 As shown, taking a power plant with a new type of power system as the target application as an example, in the process of building a simulation model, various units (such as photovoltaics, energy storage, etc.), equipment (such as inverters, PCS, etc.), and devices (such as AC / DC converters, energy storage batteries, etc.) in the power plant can be decomposed. Using multi-threading or multi-processing, one thread or process can execute the simulation program of one unit, equipment, or device to obtain the equipment model. At the same time, all units / equipment / devices of the entire power plant are decomposed into a DAG (Directed Acyclic Graph). Shuffle nodes are introduced to splice and summarize the data output by the decomposed units / equipment / devices to obtain the site model, thereby generating complete simulation data for each device of the power plant.

[0357] Each shuffle can also be executed based on a single thread or process, thereby achieving high concurrency across the entire system.

[0358] The process of breaking down the energy station into multiple units / devices / components, as well as the division of the shuffle, can be manually designed by simulation engineers using the simulation design tools mentioned in the above embodiments.

[0359] According to the energy scheduling method provided in the embodiments of this application, the target application object is split according to the actual characteristics of the target application object in order to perform local modeling and overall modeling, and obtain equipment model and site model, which helps to achieve high concurrency of the entire system and meet high computational overhead and performance requirements.

[0360] In some embodiments, prior to step 110, the method may further include:

[0361] A training sample set is constructed using the sample operating parameters of the target application object corresponding to multiple sample sub-periods as samples and the sample sub-strategies derived based on the sample operating parameters as labels. The sample operating parameters are obtained by simulation using the second simulation model based on the sample operating parameters and sample power prediction data corresponding to the previous sample sub-period. The sample operating parameters are used to characterize the real-time operating parameters of the target application object in each sample sub-period.

[0362] An energy scheduling model is constructed based on the training sample set.

[0363] In this embodiment, the categories of sample operating parameters can be the same as those of prediction operating parameters, including but not limited to: power of electrical equipment in each sub-period, power of energy storage equipment, state of charge of energy storage equipment, and power of grid incoming lines, etc.

[0364] The sample operating parameters corresponding to the current sample sub-period can be obtained by simulation modeling using the sample operating parameters and sample power prediction data corresponding to the previous sample sub-period.

[0365] Based on the sample operating parameters, sample power forecast data, sample historical data, and electricity price data corresponding to the current sample sub-period, a corresponding sample scheduling strategy can be derived, thus obtaining a set of training samples. In a similar manner, multiple sets of training samples can be obtained. These multiple sets of training samples are then input into the energy scheduling model, and the model is trained with the goal of outputting the corresponding predictive scheduling strategy, thereby obtaining the energy scheduling model.

[0366] According to the energy scheduling method provided in the embodiments of this application, during the construction and training of the energy scheduling model, simulation data and sample scheduling strategies derived from the simulation data and simulation model are used to train the energy scheduling model, so that the energy scheduling model can better adapt to the simulation data, improve the accuracy and precision of the energy scheduling model in the simulation scenario, and thus improve the accuracy of the final obtained scheduling strategy.

[0367] Figure 8 The diagram shows the scheduling effect of a park using the method of this application. Through photovoltaic power generation, energy storage for peak shaving and valley filling, orderly charging of electric vehicles, orderly discharge of mobile energy replenishment equipment and V2G equipment, the peak load of the park can be reduced by 39%, the cost per kilowatt-hour can be reduced by 22%, and the return on investment can be achieved by about 18%, which has a high scheduling effect.

[0368] This application also provides an energy scheduling method.

[0369] like Figure 9 As shown, this energy scheduling method is applied to a simulation model, and the method includes:

[0370] Step 910: When receiving the sub-policy corresponding to the sub-period in the target time period sent by the energy scheduling model, the first simulation model performs simulation deduction based on the sub-policy to obtain the sub-execution result corresponding to the sub-period.

[0371] In this embodiment, the simulation model includes a first simulation model and a second simulation model.

[0372] The sub-execution results are used to determine the simulation results corresponding to the target time period, which includes multiple sub-time periods. The simulation results are used to adjust the model parameters of the energy scheduling model in order to obtain the target scheduling strategy corresponding to the target application object.

[0373] The sub-strategy corresponding to the sub-period is predicted by the energy scheduling model based on the prediction working parameters and target data corresponding to the sub-period.

[0374] According to the energy scheduling method provided in the embodiments of this application, the energy scheduling model generates sub-strategies based on predicted short-term data, and the first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target scheduling strategy that better meets user needs. This method is applicable to complex and ever-changing scheduling scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with scheduling strategies that meet their needs as much as possible, thereby improving the user experience.

[0375] In some embodiments, before the first simulation model performs simulation deduction based on the sub-policy corresponding to a sub-time period in the target time period and obtains the sub-execution result corresponding to the sub-time period, when receiving the sub-policy corresponding to the sub-time period sent by the energy scheduling model, the method may further include:

[0376] The second simulation model is used to simulate and deduce the prediction working parameters corresponding to the previous sub-period and the power prediction data in the target data to obtain the prediction working parameters corresponding to the sub-period.

[0377] Send the prediction parameters corresponding to the sub-period to the energy scheduling model.

[0378] In some embodiments, when receiving a sub-policy corresponding to a sub-period within a target time period sent by an energy scheduling model, the first simulation model performs simulation deduction based on the sub-policy to obtain the sub-execution result corresponding to the sub-period, which may include:

[0379] If the sub-period is not the last sub-period of the target period, the first simulation model will simulate and deduce the sub-strategy corresponding to the sub-period to obtain the sub-execution result corresponding to the sub-period;

[0380] If the sub-period is the last sub-period of the target period, the first simulation model ends the simulation and, based on the sub-execution results corresponding to each sub-period, obtains the simulation results corresponding to the target period.

[0381] In some embodiments, after the first simulation model performs simulation deduction based on the sub-strategy to obtain the sub-execution result corresponding to the sub-time period, the method may further include:

[0382] Based on the sub-execution results, the first simulation model is adjusted so that the sub-strategy corresponding to the next sub-period can be simulated and deduced through the adjusted first simulation model to obtain the sub-execution results corresponding to the next sub-period.

[0383] This application also provides an energy scheduling method.

[0384] like Figure 10 As shown, the energy scheduling method is applied to an energy scheduling model, and the method includes steps 1010, 1020 and 1030.

[0385] Step 1010: Upon receiving the prediction working parameters corresponding to each sub-period in the target time period, predict the sub-strategy corresponding to each sub-period in sequence based on the prediction working parameters corresponding to each sub-period and the target data; the target time period includes multiple sub-periods.

[0386] Step 1020: Send the sub-strategies corresponding to the sub-time periods to the first simulation model in sequence, so that the first simulation model can execute the sub-strategies, obtain the sub-execution results corresponding to each sub-time period, and obtain the simulation results corresponding to the target time period based on each sub-execution result;

[0387] Step 1030: After obtaining the simulation results corresponding to the target time period and adjusting the model parameters of the energy scheduling model, re-execute the step of predicting the sub-strategy corresponding to each sub-time period based on the prediction working parameters and target data of each sub-time period when the prediction working parameters of each sub-time period are received, until the preset stopping condition is reached; so as to filter the target scheduling strategy corresponding to the target application object according to the obtained simulation results.

[0388] According to the energy scheduling method provided in the embodiments of this application, the energy scheduling model generates sub-strategies based on predicted short-term data, and the first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target scheduling strategy that better meets user needs. This method is applicable to complex and ever-changing scheduling scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with scheduling strategies that meet their needs as much as possible, thereby improving the user experience.

[0389] In some embodiments, adjusting the model parameters of the energy dispatch model may include:

[0390] The model parameters of the energy dispatch model are adjusted with the goal of obtaining the optimal value of the simulation results.

[0391] In some embodiments, reaching a preset stopping condition may include:

[0392] The number of adjustments exceeds the first threshold.

[0393] In some embodiments, receiving prediction parameters corresponding to each sub-time period in the target time period may include:

[0394] The system receives the prediction parameters corresponding to the sub-period sent by the second simulation model. The prediction parameters corresponding to the sub-period are obtained by the second simulation model through simulation and deduction of the prediction parameters corresponding to the previous sub-period and the power prediction data in the target data.

[0395] In some embodiments, after selecting the target scheduling strategy corresponding to the target application object based on the obtained simulation results, the method may further include:

[0396] Update the actual values ​​of the model parameters to the model parameters corresponding to the target scheduling strategy.

[0397] This application also provides an energy scheduling method.

[0398] like Figure 11 As shown, this energy dispatching method is applied to an energy dispatching system, and the method includes:

[0399] Step 1110: The energy scheduling model sequentially predicts the prediction working parameters and target data corresponding to each sub-period in the target time period to obtain the sub-strategy corresponding to each sub-period; the target time period includes multiple sub-periods.

[0400] Step 1120: The first simulation model in the simulation model performs simulation deduction on each sub-strategy in sequence to obtain the sub-execution results corresponding to each sub-time period;

[0401] Step 1130: Based on the simulation results obtained from the sub-execution results corresponding to each sub-period in the target time period, the controller adjusts the model parameters of the energy scheduling model;

[0402] Step 1140: The adjusted energy scheduling model re-executes the steps of predicting the prediction working parameters and target data corresponding to each sub-period in the target time period in turn, and obtaining the sub-strategy corresponding to each sub-period, until the preset stopping condition is reached.

[0403] Step 1150: The controller selects the target scheduling strategy corresponding to the target application object based on the simulation results.

[0404] In this embodiment, the energy scheduling system includes a controller, an energy scheduling model, and a simulation model deployed on a simulation service cluster, with the simulation model and the energy scheduling model connected together.

[0405] The simulation model includes a first simulation model and a second simulation model.

[0406] The controller is connected to the energy dispatch model and the simulation model respectively, and is used to execute steps 110 to 150.

[0407] The specific execution steps of the energy dispatch system have been described in the above embodiments, and will not be repeated here.

[0408] According to the energy scheduling method provided in the embodiments of this application, the energy scheduling model generates sub-strategies based on predicted short-term data, and the first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target scheduling strategy that better meets user needs. This method is applicable to complex and ever-changing scheduling scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with scheduling strategies that meet their needs as much as possible, thereby improving the user experience.

[0409] In some embodiments, before the energy scheduling model sequentially predicts the prediction working parameters and target data corresponding to each sub-period in the target time period to obtain the sub-strategy corresponding to each sub-period, the method may further include:

[0410] The second simulation model in the simulation model performs simulation and deduction on the predicted operating parameters corresponding to the previous sub-period and the power prediction data in the target data to obtain the predicted operating parameters corresponding to the sub-period, and sends the predicted operating parameters corresponding to the sub-period to the energy dispatch model.

[0411] In some embodiments, the controller may filter the target scheduling strategy corresponding to the target application object based on the simulation results, which may include:

[0412] Based on the simulation results, calculate the evaluation score corresponding to the simulation results from at least one dimension;

[0413] The scheduling strategy corresponding to the simulation result with the highest evaluation score is determined as the target scheduling strategy.

[0414] The energy scheduling method provided in this application can be executed by an energy scheduling device. This application uses an energy scheduling device executing the energy scheduling method as an example to illustrate the energy scheduling device provided in this application.

[0415] This application also provides an energy dispatching device.

[0416] like Figure 12 As shown, the energy dispatching device includes: a first processing module 1210, a second processing module 1220, a third processing module 1230, a fourth processing module 1240, and a fifth processing module 1250.

[0417] The first processing module 1210 is used to obtain the current sub-period and the prediction working parameters and target data corresponding to each remaining sub-period in the target time period;

[0418] The second processing module 1220 is used to sequentially input the prediction working parameters and target data corresponding to each sub-period into the energy scheduling model to obtain the sub-strategy corresponding to each sub-period.

[0419] The third processing module 1230 is used to sequentially input the sub-strategies corresponding to the sub-time periods into the first simulation model to obtain the sub-execution results corresponding to each sub-time period.

[0420] The fourth processing module 1240 is used to adjust the model parameters of the energy scheduling model based on the simulation results obtained from the sub-execution results corresponding to each sub-period in the target time period, and to re-execute the steps of obtaining the prediction working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period until the preset stopping condition is reached.

[0421] The fifth processing module 1250 is used to filter and obtain the target scheduling strategy corresponding to the target application object based on the simulation results.

[0422] According to the energy scheduling device provided in the embodiments of this application, the energy scheduling model generates sub-strategies based on predicted short-term data, and the first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target scheduling strategy that better meets user needs. This device is applicable to complex and ever-changing scheduling scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with scheduling strategies that meet their needs as much as possible, thereby improving the user experience.

[0423] In some embodiments, the first processing module 1210 is configured to:

[0424] Based on the forecasting parameters corresponding to the previous sub-period and the power forecasting data in the target data, the forecasting parameters corresponding to the current sub-period are obtained.

[0425] In some embodiments, the first processing module 1210 is configured to:

[0426] The prediction parameters corresponding to the previous sub-period and the power prediction data in the target data are input into the second simulation model to obtain the prediction parameters corresponding to the sub-period.

[0427] In some embodiments, the first processing module 1210 is configured to:

[0428] The second simulation model is adjusted based on the sub-execution result corresponding to the previous sub-period.

[0429] The prediction parameters corresponding to the previous sub-period and the power prediction data in the target data are input into the adjusted second simulation model to obtain the prediction parameters corresponding to the sub-period.

[0430] In some embodiments, the second processing module 1220 is configured to:

[0431] For each sub-period, that sub-period is taken as the current sub-period, and the first simulation model is adjusted according to the sub-execution result corresponding to the current sub-period;

[0432] The sub-strategy corresponding to the next sub-period of the current sub-period is input into the adjusted first simulation model to obtain the sub-execution result corresponding to the next sub-period of the current sub-period.

[0433] In some embodiments, the second processing module 1220 is configured to:

[0434] The sub-strategy corresponding to the sub-period is input into the first simulation model so that at least one target device model in the first simulation model can execute the sub-strategy to obtain the sub-execution result corresponding to the sub-period. The sub-execution result includes the simulation working parameters of the target device model and the device model associated with the target device model after the sub-execution result is executed. The simulation working parameters are used to adjust the first simulation model so that the adjusted first simulation model can simulate and deduce the sub-execution result corresponding to the next sub-period of the sub-period.

[0435] In some embodiments, the second processing module 1220 is configured to:

[0436] The second simulation model is used to simulate and deduce the prediction working parameters corresponding to the previous sub-period of each sub-period until the prediction working parameters corresponding to each sub-period included in the target period are obtained.

[0437] The energy scheduling model processes the predicted working parameters and target data for each sub-period within the target time period to obtain the scheduling strategy corresponding to the target time period. The scheduling strategy includes the sub-strategies corresponding to each sub-period.

[0438] In some embodiments, the fourth processing module 1240 is configured to:

[0439] Perform the following operations sequentially according to the order of the sub-time periods:

[0440] S1. Initialize the current sub-time period;

[0441] S2. Obtain the prediction parameters and target data corresponding to the current sub-period;

[0442] S3. Input the prediction working parameters and target data corresponding to the current sub-period into the energy scheduling model to obtain the sub-strategy corresponding to the current sub-period.

[0443] S4. Input the sub-strategy corresponding to the current sub-period into the first simulation model to obtain the sub-execution result corresponding to the sub-strategy corresponding to the current sub-period.

[0444] S5. If the current sub-period is not the last sub-period of the target period, then return to execute S1;

[0445] S6. If the current sub-period is the last sub-period of the target period, then terminate the operation.

[0446] In some embodiments, the fourth processing module 1240 is configured to:

[0447] If the number of adjustments is less than or equal to the first threshold, the steps of obtaining the current sub-period and the prediction working parameters and target data corresponding to each remaining sub-period in the target period are re-executed.

[0448] If the number of adjustments exceeds the first threshold, the steps of obtaining the current sub-period and the prediction working parameters and target data corresponding to each remaining sub-period in the target time period are stopped.

[0449] In some embodiments, the fourth processing module 1240 is configured to:

[0450] The model parameters of the energy dispatch model are adjusted with the goal of obtaining the optimal value of the simulation results;

[0451] or,

[0452] A subset of parameters from the energy dispatch model were randomly selected and determined as the parameters to be adjusted.

[0453] Adjust the parameter to be adjusted so that the adjusted parameter is different from the parameter after each previous adjustment;

[0454] or,

[0455] The model parameters of the energy scheduling model are adjusted using a tuner.

[0456] or,

[0457] The model parameters of the energy dispatch model are adjusted using a loss function; the loss function is constructed based on simulation results and target simulation results.

[0458] In some embodiments, the fifth processing module 1250 is configured to:

[0459] Based on the simulation results, calculate the evaluation score corresponding to the simulation results from at least one dimension;

[0460] The scheduling strategy corresponding to the simulation result with the highest evaluation score is determined as the target scheduling strategy.

[0461] In some embodiments, the apparatus further includes a fifteenth processing module for:

[0462] Before obtaining the prediction working parameters and target data corresponding to the current sub-time period and the remaining sub-time periods in the target time period, a first simulation model and a second simulation model are constructed based on at least one of the following: the equipment information of the target application object, the local control strategy corresponding to the equipment included in the target application object, the power topology information of the target application object, and the overall coordination control strategy of the target application object.

[0463] In some embodiments, the device further includes a sixteenth processing module for:

[0464] Before obtaining the prediction working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period, a training sample set is constructed using the sample working parameters of the target application object in multiple sample sub-periods as samples and the sample sub-strategies derived based on the sample working parameters as labels. The sample working parameters are obtained by simulation using the second simulation model based on the sample working parameters and sample power prediction data corresponding to the previous sample sub-period. The sample working parameters are used to characterize the real-time working parameters of the target application object in each sample sub-period.

[0465] An energy scheduling model is constructed based on the training sample set.

[0466] In some embodiments, the device further includes a seventeenth processing module for:

[0467] After selecting the target scheduling strategy corresponding to the target application object based on the simulation results, the model parameters corresponding to the target scheduling strategy are determined as the actual model parameters of the energy scheduling model in the target time period.

[0468] This application also provides an energy dispatching device.

[0469] like Figure 13 As shown, the energy dispatching device is applied to a simulation model, which includes a first simulation model; the device includes:

[0470] The sixth processing module 1310 is used to enable the first simulation model to perform simulation deduction based on the sub-strategy in the sub-period of the target time period when receiving the sub-strategy sent by the energy scheduling model, so as to obtain the sub-execution result corresponding to the sub-period; wherein, the sub-execution result is used to determine the simulation result corresponding to the target time period, the target time period includes multiple sub-periods, and the simulation result is used to adjust the model parameters of the energy scheduling model in order to obtain the target scheduling strategy corresponding to the target application object;

[0471] The sub-strategy corresponding to the sub-period is predicted by the energy scheduling model based on the prediction working parameters and target data corresponding to the sub-period.

[0472] According to the energy scheduling device provided in the embodiments of this application, the energy scheduling model generates sub-strategies based on predicted short-term data, and the first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target scheduling strategy that better meets user needs. This device is applicable to complex and ever-changing scheduling scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with scheduling strategies that meet their needs as much as possible, thereby improving the user experience.

[0473] In some embodiments, the simulation model further includes a second simulation model; the apparatus further includes an eighteenth processing module, configured to:

[0474] When receiving the sub-strategy corresponding to the sub-period in the target time period sent by the energy dispatch model, before the first simulation model performs simulation deduction based on the sub-strategy and obtains the sub-execution result corresponding to the sub-period, the second simulation model performs simulation deduction on the prediction working parameters corresponding to the previous sub-period and the power prediction data in the target data to obtain the prediction working parameters corresponding to the sub-period.

[0475] The second simulation model sends the prediction parameters corresponding to the sub-period to the energy scheduling model.

[0476] In some embodiments, the sixth processing module 1310 is configured to:

[0477] If the sub-period is not the last sub-period of the target period, the first simulation model will simulate and deduce the sub-strategy corresponding to the sub-period to obtain the sub-execution result corresponding to the sub-period;

[0478] If the sub-period is the last sub-period of the target period, the first simulation model ends the simulation and, based on the sub-execution results corresponding to each sub-period, obtains the simulation results corresponding to the target period.

[0479] In some embodiments, the apparatus further includes a nineteenth processing module for:

[0480] After the first simulation model performs simulation deduction based on the sub-strategy and obtains the sub-execution result corresponding to the sub-period, the first simulation model is adjusted according to the sub-execution result. The adjusted first simulation model is then used to perform simulation deduction on the sub-strategy corresponding to the next sub-period and obtain the sub-execution result corresponding to the next sub-period.

[0481] This application also provides an energy dispatching device.

[0482] like Figure 14 As shown, this energy dispatching device is applied to an energy dispatching model. The device includes:

[0483] The seventh processing module 1410 is used to enable the energy scheduling model to predict the sub-strategy corresponding to each sub-period based on the prediction working parameters corresponding to each sub-period in the target time period and the target data, when it receives the prediction working parameters corresponding to each sub-period in the target time period; the target time period includes multiple sub-periods.

[0484] The eighth processing module 1420 is used to enable the energy scheduling model to send the sub-strategies corresponding to the sub-time periods to the first simulation model in sequence, so that the first simulation model can execute the sub-strategies, obtain the sub-execution results corresponding to each sub-time period, and obtain the simulation results corresponding to the target time period based on each sub-execution result;

[0485] The ninth processing module 1430 is used to, after obtaining the simulation results corresponding to the target time period and adjusting the model parameters of the energy scheduling model, cause the energy scheduling model to re-execute the steps of predicting the sub-strategies corresponding to each sub-time period based on the prediction working parameters corresponding to each sub-time period and the target data, until a preset stopping condition is reached; so as to filter the target scheduling strategy corresponding to the target application object according to the obtained simulation results.

[0486] According to the energy scheduling device provided in the embodiments of this application, the energy scheduling model generates sub-strategies based on predicted short-term data, and the first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target scheduling strategy that better meets user needs. This device is applicable to complex and ever-changing scheduling scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with scheduling strategies that meet their needs as much as possible, thereby improving the user experience.

[0487] In some embodiments, the ninth processing module 1430 is configured to:

[0488] The model parameters of the energy dispatch model are adjusted with the goal of obtaining the optimal value of the simulation results.

[0489] In some embodiments, the seventh processing module 1410 is configured to:

[0490] The energy dispatch model receives the predicted operating parameters corresponding to the sub-period sent by the second simulation model. The predicted operating parameters corresponding to the sub-period are obtained by the second simulation model through simulation and deduction of the predicted operating parameters corresponding to the previous sub-period and the power prediction data in the target data.

[0491] In some embodiments, the device further includes a twentieth processing module for:

[0492] After selecting the target scheduling strategy corresponding to the target application object based on the obtained simulation results, the actual values ​​of the model parameters are updated to the model parameters corresponding to the target scheduling strategy.

[0493] This application also provides an energy dispatching device.

[0494] like Figure 15 As shown, this energy dispatching device is applied to an energy dispatching system, which includes a controller, an energy dispatching model, and a simulation model deployed in a simulation service cluster. The simulation model and the energy dispatching model are connected. The device includes:

[0495] The tenth processing module 1510 is used to enable the energy scheduling model to predict the prediction working parameters and target data corresponding to each sub-period in the target time period in sequence, so as to obtain the sub-strategy corresponding to each sub-period; the target time period includes multiple sub-periods;

[0496] The eleventh processing module 1520 is used to enable the first simulation model in the simulation model to perform simulation and deduction on each sub-strategy in sequence, and obtain the sub-execution results corresponding to each sub-time period.

[0497] The twelfth processing module 1530 is used to adjust the model parameters of the energy scheduling model by the controller when the simulation results are obtained based on the sub-execution results corresponding to each sub-time period in the target time period.

[0498] The thirteenth processing module 1540 is used to enable the adjusted energy scheduling model to re-execute the steps of predicting the prediction working parameters and target data corresponding to each sub-period in the target time period, and obtaining the sub-strategy corresponding to each sub-period until the preset stopping condition is reached.

[0499] The fourteenth processing module 1550 is used to enable the controller to filter and obtain the target scheduling strategy corresponding to the target application object based on the simulation results.

[0500] According to the energy scheduling device provided in the embodiments of this application, the energy scheduling model generates sub-strategies based on predicted short-term data, and the first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target scheduling strategy that better meets user needs. This device is applicable to complex and ever-changing scheduling scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with scheduling strategies that meet their needs as much as possible, thereby improving the user experience.

[0501] In some embodiments, the apparatus further includes a twenty-first processing module for:

[0502] Before the energy dispatch model sequentially predicts the predicted operating parameters and target data corresponding to each sub-period in the target time period to obtain the sub-strategy corresponding to each sub-period, the second simulation model in the simulation model performs simulation and deduction on the predicted operating parameters and power prediction data in the target data corresponding to the previous sub-period of the sub-period to obtain the predicted operating parameters corresponding to the sub-period, and sends the predicted operating parameters corresponding to the sub-period to the energy dispatch model.

[0503] In some embodiments, the fourteenth processing module 1550 is configured to:

[0504] Based on the simulation results, calculate the evaluation score corresponding to the simulation results from at least one dimension;

[0505] The scheduling strategy corresponding to the simulation result with the highest evaluation score is determined as the target scheduling strategy.

[0506] The energy dispatching device in this application embodiment can be an electronic device or a component within an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a mobile phone, tablet computer, laptop computer, PDA, in-vehicle electronic device, mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc. It can also be a server, network attached storage (NAS), personal computer (PC), television set (TV), ATM, or self-service machine, etc. This application embodiment does not specifically limit the device.

[0507] The energy dispatching device in this application embodiment can be a device with an operating system. This operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit the specific operating system used.

[0508] The energy scheduling device provided in this application embodiment can achieve… Figures 1 to 11 The various processes implemented in the method implementation examples will not be described again here to avoid repetition.

[0509] This application also provides an energy dispatching system.

[0510] like Figure 3 As shown, the energy dispatch system includes: a controller, a simulation model, an energy dispatch model, and a controller.

[0511] The simulation model is deployed in the simulation service cluster. The controller is connected to the simulation service cluster and controls the target computing node in the simulation service cluster to enable the model in the simulation model corresponding to the target computing node to perform simulation and deduction.

[0512] In some embodiments, the connection may include a wired connection and a wireless connection.

[0513] In some embodiments, the simulation model includes at least one of a first simulation model and a second simulation model.

[0514] The energy scheduling model is connected to both the simulation model and the controller.

[0515] The controller is used to execute the energy scheduling method as described in the above embodiments.

[0516] Continue to refer to Figure 3 In some embodiments, the energy scheduling system includes: an online simulation subsystem, an energy scheduling strategy deduction subsystem, and an algorithm training system.

[0517] The online simulation subsystem includes a simulation model.

[0518] The energy scheduling strategy deduction subsystem consists of three parts: "algorithm execution module", "effect evaluation module" and "algorithm optimization module".

[0519] The algorithm execution module is used to receive the simulation data (i.e., predicted operating parameters) continuously input from the online simulation subsystem, execute the scheduling strategy generated by the energy scheduling algorithm, and calculate the simulation results of the scheduling strategy.

[0520] The energy scheduling model can be deployed in the energy scheduling strategy derivation subsystem to generate scheduling strategies based on the received predictive working parameters.

[0521] The performance evaluation module evaluates the effectiveness of the scheduling strategy on a daily basis based on the daily simulation results output by the algorithm execution module. For example, it compares the scheduling benefits of all energy scheduling algorithms output by the algorithm optimization module in this simulation and selects the scheduling algorithm with the highest benefit.

[0522] The algorithm optimization module is used to locally fine-tune the configuration parameters of the energy scheduling algorithm based on the performance evaluation results of the previous round, until the model parameters that best meet the user's needs are found. These model parameters can then be used as the optimal configuration parameters identified in this iteration. This module can be implemented using commonly used parameter tuning tuners in the industry, such as random tuner, grid tuner, and GA tuner based on genetic algorithms.

[0523] An algorithm training system may include modules such as a "training dataset", an "algorithm training engine", and an "algorithm deployment engine".

[0524] The training dataset is used to maintain the set of training data for the energy scheduling algorithm. The governance tools for the training dataset can adopt industry-standard solutions.

[0525] The algorithm training engine is used to train and fine-tune the model parameters of the energy scheduling model.

[0526] The algorithm deployment engine is used to deploy the optimized energy scheduling algorithm. For example, Docker technology can be used to encapsulate and deploy the algorithm model as an algorithm model; or other mainstream algorithm OTA solutions can be used, etc. This application does not limit this.

[0527] After generating the optimal day-ahead and mid-day scheduling strategies, the energy dispatch system automatically imports these strategies into the algorithm training engine. The engine combines the newly added scheduling strategies with simulation data, along with essential data such as the current, voltage, and power of the microgrid's source-grid-load-storage equipment on the day, into the training or validation set of the scheduling strategy algorithm. This retrains the algorithm to generate a globally optimal model. Then, the energy dispatch algorithm model of the energy dispatch system is remotely upgraded via the OTA (Over-The-Air) function.

[0528] The energy dispatching system provided in this application embodiment generates sub-strategies based on predicted short-term data by an energy dispatching model. A first simulation model executes the sub-strategies to generate sub-execution results. By adopting a dual-loop mode, the model parameters are optimized with the goal of selecting a target dispatching strategy that better meets user needs. This system is applicable to complex and ever-changing dispatching scenarios caused by seasonal changes, weather changes, and changes in user energy consumption scenarios. It has high real-time performance, meets the requirements of real-time or near-real-time updates, and continuously provides users with dispatching strategies that meet their needs as much as possible, thereby improving the user experience.

[0529] The energy dispatching method provided in this application can be used for energy dispatching of energy storage devices, power consumption devices, and power generation devices.

[0530] The energy storage device includes one or more battery clusters to increase its voltage and capacity. A battery cluster may include multiple battery units connected in series via a busbar to increase the voltage of the energy storage device. When the energy storage device includes multiple battery clusters, the clusters are connected in parallel to increase the capacity of the energy storage device.

[0531] Energy storage devices can be used in energy storage power stations, wind power generation systems, solar power generation systems, mobile power systems, or temporary power supply systems. Energy storage devices can store electrical energy as needed and output it when appropriate. For example, an energy storage device can store electrical energy during off-peak hours and provide power to relevant users or electrical devices during peak hours. The energy storage system provided in this application embodiment can be any power system that requires energy storage devices.

[0532] In some embodiments, the energy storage device is an energy storage container or an energy storage cabinet.

[0533] In some embodiments, the energy storage device may include a cabinet and one or more battery clusters housed within the cabinet.

[0534] In some embodiments, the energy storage device may include individual battery cells.

[0535] In this embodiment of the application, the battery cell can be a secondary battery, which refers to a battery cell that can be recharged to activate the active materials and continue to be used after the battery cell has been discharged.

[0536] The battery cell can be a lithium-ion battery, sodium-ion battery, sodium-lithium-ion battery, lithium metal battery, sodium metal battery, lithium-sulfur battery, magnesium-ion battery, nickel-metal hydride battery, nickel-cadmium battery, lead-acid battery, etc., and the embodiments of this application are not limited to this.

[0537] In some embodiments, the energy storage device may include a battery device.

[0538] The battery apparatus mentioned in the embodiments of this application may include one or more battery cell assemblies for providing voltage and capacity. A battery cell assembly may include multiple battery cells connected in series, parallel, or mixed connections via a busbar.

[0539] In some embodiments, a battery cell assembly is typically formed by arranging multiple battery cells.

[0540] As an example, a battery cell assembly can be a battery module, which is formed by arranging and fixing multiple battery cells together to form an independent module. As another example, a battery module can be formed by bundling multiple battery cells together with cable ties.

[0541] In some embodiments, the battery device may be a battery pack, which includes a housing and one or more individual battery cells housed within the housing.

[0542] Electrical devices can include, but are not limited to, mobile phones, tablets, laptops, electric toys, power tools, electric vehicles, electric cars, ships, and spacecraft. Among them, electric toys can include stationary or mobile electric toys, such as game consoles, electric car toys, electric ship toys, and electric airplane toys, etc., and spacecraft can include airplanes, rockets, space shuttles, and spacecraft, etc.

[0543] The power generation device can be a traditional power generation device or a new energy power generation device. It includes, but is not limited to: photovoltaic power generation devices, wind power generation devices, tidal power generation devices, and other power generation devices.

[0544] In some embodiments, such as Figure 16As shown, this application embodiment also provides an electronic device 1600, including a processor 1601, a memory 1602, and a computer program stored in the memory 1602 and executable on the processor 1601. When the program is executed by the processor 1601, it implements the various processes of the above-described energy dispatching method embodiment and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0545] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0546] This application also provides a non-transitory computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described energy scheduling method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0547] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0548] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described energy scheduling method.

[0549] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0550] This application embodiment also provides a chip, which includes a processor and a communication interface. The communication interface is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above-described energy scheduling method embodiments and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0551] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0552] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0553] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0554] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0555] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0556] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

Claims

1. An energy scheduling method, characterized in that, include: Obtain the current sub-period and the prediction parameters and target data for each remaining sub-period within the target time period; The prediction working parameters and target data corresponding to each sub-time period are sequentially input into the energy scheduling model to obtain the sub-strategy corresponding to each sub-time period. The sub-strategies corresponding to the sub-time periods are sequentially input into the first simulation model to obtain the sub-execution results corresponding to each sub-time period. When the simulation results are obtained based on the sub-execution results corresponding to each sub-period in the target time period, the model parameters of the energy scheduling model are adjusted, and the steps of obtaining the prediction working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period are re-executed until the preset stopping condition is reached. The target scheduling strategy corresponding to the target application object is obtained by filtering based on the simulation results.

2. The energy dispatching method according to claim 1, characterized in that, The process of obtaining the prediction parameters and target data corresponding to the current sub-period and the remaining sub-periods within the target time period includes: Based on the prediction parameters corresponding to the previous sub-period of the current sub-period and the power prediction data in the target data, the prediction parameters corresponding to the current sub-period are obtained.

3. The energy dispatching method according to claim 2, characterized in that, The process of obtaining the prediction parameters corresponding to the current sub-period based on the prediction parameters corresponding to the previous sub-period and the power prediction data in the target data includes: The prediction parameters corresponding to the previous sub-period of the current sub-period and the power prediction data in the target data are input into the second simulation model to obtain the prediction parameters corresponding to the sub-period.

4. The energy dispatching method according to claim 2 or 3, characterized in that, The process of obtaining the prediction parameters corresponding to the current sub-period based on the prediction parameters corresponding to the previous sub-period and the power prediction data in the target data includes: The second simulation model is adjusted based on the sub-execution result corresponding to the previous sub-period of the current sub-period. The prediction parameters corresponding to the previous sub-period and the power prediction data in the target data are input into the adjusted second simulation model to obtain the prediction parameters corresponding to the current sub-period.

5. The energy dispatching method according to any one of claims 1-4, characterized in that, The step of sequentially inputting the prediction working parameters and target data corresponding to each of the sub-time periods into the energy scheduling model to obtain the sub-strategy corresponding to each of the sub-time periods includes: For each sub-period, the sub-period is taken as the current sub-period, and the first simulation model is adjusted according to the sub-execution result corresponding to the current sub-period; The sub-strategy corresponding to the next sub-period of the current sub-period is input into the adjusted first simulation model to obtain the sub-execution result corresponding to the next sub-period of the current sub-period.

6. The energy scheduling method according to any one of claims 1-5, characterized in that, The step of sequentially inputting the prediction working parameters and target data corresponding to each of the sub-time periods into the energy scheduling model to obtain the sub-strategy corresponding to each of the sub-time periods includes: The sub-strategy corresponding to the sub-time period is input into the first simulation model so that at least one target device model in the first simulation model can execute the sub-strategy to obtain the sub-execution result corresponding to the sub-time period. The sub-execution result includes the simulation working parameters of the target device model and the device model associated with the target device model after executing the sub-execution result. The simulation working parameters are used to adjust the first simulation model so that the adjusted first simulation model can simulate and deduce the sub-execution result corresponding to the next sub-time period of the sub-time period.

7. The energy dispatching method according to any one of claims 1-6, characterized in that, The step of sequentially inputting the prediction working parameters and target data corresponding to each of the sub-time periods into the energy scheduling model to obtain the sub-strategy corresponding to each of the sub-time periods includes: The prediction parameters corresponding to the previous sub-period of each sub-period are simulated and deduced using the second simulation model until the prediction parameters corresponding to each sub-period included in the target period are obtained. The energy scheduling model processes the prediction parameters and target data for each sub-period within the target time period to obtain the scheduling strategy corresponding to the target time period. The scheduling strategy includes sub-strategies corresponding to each sub-period.

8. The energy dispatching method according to any one of claims 1-7, characterized in that, The step of adjusting the model parameters of the energy scheduling model and re-executing the steps of obtaining the prediction working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period, based on the simulation results obtained from the sub-execution results corresponding to each sub-period in the target time period, includes: The following operations are performed sequentially according to the order of the sub-time periods: S1. Initialize the current sub-time period; S2. Obtain the prediction working parameters and target data corresponding to the current sub-time period; S3. Input the prediction working parameters and target data corresponding to the current sub-period into the energy scheduling model to obtain the sub-strategy corresponding to the current sub-period; S4. Input the sub-strategy corresponding to the current sub-period into the first simulation model to obtain the sub-execution result corresponding to the sub-strategy corresponding to the current sub-period; S5. If the current sub-period is not the last sub-period of the target period, then return to execute S1; S6. If the current sub-time period is the last sub-time period of the target time period, then terminate the operation.

9. The energy dispatching method according to any one of claims 1-8, characterized in that, When the simulation results are obtained based on the sub-execution results corresponding to each sub-time period in the target time period, the model parameters of the energy scheduling model are adjusted, and the steps of obtaining the prediction working parameters and target data corresponding to the current sub-time period and the remaining sub-time periods in the target time period are re-executed until the preset stopping condition is reached, including: If the number of adjustments is less than or equal to the first threshold, the step of obtaining the current sub-period and the prediction working parameters and target data corresponding to each remaining sub-period in the target period is re-executed. If the number of adjustments exceeds the first threshold, the step of obtaining the prediction working parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period shall be stopped.

10. The energy dispatching method according to any one of claims 1-9, characterized in that, The adjustment of the model parameters of the energy scheduling model includes at least one of the following methods: The model parameters of the energy dispatch model are adjusted with the goal of obtaining the optimal value of the simulation results; or, Randomly select some parameters of the energy dispatch model to determine the parameters to be adjusted; Adjust the parameter to be adjusted so that the adjusted parameter is different from the parameter after each previous adjustment; or, The model parameters of the energy scheduling model are adjusted using a tuner. or, The model parameters of the energy scheduling model are adjusted using a loss function; the loss function is constructed based on the simulation results and the target simulation results.

11. The energy dispatching method according to any one of claims 1-10, characterized in that, The step of obtaining the target scheduling strategy corresponding to the target application object based on the simulation results includes: Based on the simulation results, calculate the evaluation score corresponding to the simulation results from at least one dimension; The scheduling strategy corresponding to the simulation result with the highest evaluation score is determined as the target scheduling strategy.

12. The energy dispatching method according to claim 11, characterized in that, The at least one dimension includes the total revenue corresponding to the target time period.

13. The energy dispatching method according to any one of claims 1-12, characterized in that, Before obtaining the prediction parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period, the method further includes: Based on at least one of the following: the device information of the target application object, the local control strategy corresponding to the device included in the target application object, the power topology information of the target application object, and the overall coordination control strategy of the target application object, the first simulation model and the second simulation model are constructed.

14. The energy dispatching method according to any one of claims 1-13, characterized in that, Before obtaining the prediction parameters and target data corresponding to the current sub-period and the remaining sub-periods in the target time period, the method further includes: A training sample set is constructed using the sample operating parameters of the target application object corresponding to multiple sample sub-time periods as samples and the sample sub-strategies derived based on the sample operating parameters as labels. The sample operating parameters are obtained by simulation using a second simulation model based on the sample operating parameters and sample power prediction data corresponding to the previous sample sub-time period. The sample operating parameters are used to characterize the real-time operating parameters of the target application object in each of the sample sub-time periods. The energy scheduling model is constructed based on the training sample set.

15. The energy dispatching method according to any one of claims 1-13, characterized in that, After obtaining the target scheduling strategy corresponding to the target application object based on the simulation results, the method further includes: The model parameters corresponding to the target scheduling strategy are determined as the actual model parameters of the energy scheduling model during the target time period.

16. An energy scheduling method, characterized in that, The method is applied to a simulation model, the simulation model including a first simulation model; the method includes: When receiving the sub-policy corresponding to a sub-time period within the target time period sent by the energy scheduling model, the first simulation model performs simulation deduction based on the sub-policy to obtain the sub-execution result corresponding to the sub-time period; wherein... The sub-execution results are used to determine the simulation results corresponding to the target time period, which includes multiple sub-time periods. The simulation results are used to adjust the model parameters of the energy scheduling model to obtain the target scheduling strategy corresponding to the target application object. The sub-strategy corresponding to the sub-period is predicted by the energy scheduling model based on the prediction working parameters and target data corresponding to the sub-period.

17. The energy dispatching method according to claim 16, characterized in that, The simulation model further includes a second simulation model; before the first simulation model performs simulation deduction based on the sub-strategy in the target time period sent by the energy scheduling model to obtain the sub-execution result corresponding to the sub-time period, the method further includes: The second simulation model is used to simulate and deduce the prediction working parameters corresponding to the previous sub-period of the sub-period and the power prediction data in the target data to obtain the prediction working parameters corresponding to the sub-period. Send the prediction parameters corresponding to the sub-period to the energy scheduling model.

18. The energy dispatching method according to claim 16 or 17, characterized in that, When receiving the sub-policy corresponding to a sub-period in the target time period sent by the energy scheduling model, the first simulation model performs simulation deduction based on the sub-policy to obtain the sub-execution result corresponding to the sub-period, including: If the sub-period is not the last sub-period of the target period, then the first simulation model performs simulation and deduction on the sub-strategy corresponding to the sub-period to obtain the sub-execution result corresponding to the sub-period; If the sub-period is the last sub-period of the target period, then the first simulation model ends the simulation and, based on the sub-execution results corresponding to each sub-period, obtains the simulation results corresponding to the target period.

19. The energy dispatching method according to any one of claims 16-18, characterized in that, After the first simulation model performs simulation deduction based on the sub-strategy to obtain the sub-execution result corresponding to the sub-time period, the method further includes: Based on the sub-execution result, the first simulation model is adjusted so that the sub-strategy corresponding to the next sub-period of the sub-period can be simulated and deduced through the adjusted first simulation model to obtain the sub-execution result corresponding to the next sub-period.

20. An energy scheduling method, characterized in that, Applied to energy scheduling models, the method includes: Upon receiving the prediction parameters corresponding to each sub-period within the target time period, the sub-strategy corresponding to each sub-period is predicted sequentially based on the prediction parameters corresponding to each sub-period and the target data; the target time period includes multiple sub-periods. The sub-strategies corresponding to the sub-time periods are sequentially sent to the first simulation model for the first simulation model to execute the sub-strategies, obtain the sub-execution results corresponding to each sub-time period, and obtain the simulation results corresponding to the target time period based on each sub-execution result; After obtaining the simulation results corresponding to the target time period and adjusting the model parameters of the energy scheduling model, the step of predicting the sub-strategy corresponding to each sub-time period based on the prediction working parameters corresponding to each sub-time period and the target data is re-executed when the prediction working parameters corresponding to each sub-time period in the target time period are received, until the preset stopping condition is reached; so as to filter the target scheduling strategy corresponding to the target application object according to the obtained simulation results.

21. The energy dispatching method according to claim 20, characterized in that, The adjustment of the model parameters of the energy scheduling model includes: The model parameters of the energy dispatch model are adjusted with the goal of obtaining the optimal value of the simulation results.

22. The energy dispatching method according to claim 20 or 21, characterized in that, The conditions for reaching the preset stop include: The number of adjustments exceeds the first threshold.

23. The energy dispatching method according to any one of claims 20-22, characterized in that, The received prediction parameters corresponding to each sub-time period in the target time period include: The system receives the prediction working parameters corresponding to the sub-time period sent by the second simulation model. The prediction working parameters corresponding to the sub-time period are obtained by the second simulation model through simulation and deduction of the prediction working parameters corresponding to the previous sub-time period and the power prediction data in the target data.

24. The energy dispatching method according to any one of claims 20-23, characterized in that, After obtaining the target scheduling strategy corresponding to the target application object based on the acquired simulation results, the method further includes: The actual values ​​of the model parameters are updated to the model parameters corresponding to the target scheduling strategy.

25. An energy dispatching method, characterized in that, The method is applied to an energy dispatching system, which includes a controller, an energy dispatching model, and a simulation model deployed on a simulation service cluster, wherein the simulation model and the energy dispatching model are connected; the method includes: The energy scheduling model sequentially predicts the prediction working parameters and target data corresponding to each sub-period in the target time period to obtain the sub-strategy corresponding to each sub-period; the target time period includes multiple sub-periods. In the simulation model, the first simulation model sequentially simulates and deduces each of the sub-strategies to obtain the sub-execution results corresponding to each of the sub-time periods; When simulation results are obtained based on the sub-execution results corresponding to each sub-time period in the target time period, the controller adjusts the model parameters of the energy scheduling model; The adjusted energy scheduling model re-executes the steps of predicting the prediction working parameters and target data corresponding to each sub-time period in the target time period in turn to obtain the sub-strategy corresponding to each sub-time period, until the preset stopping condition is reached; The controller selects the target scheduling strategy corresponding to the target application object based on the simulation results.

26. The energy dispatching method according to claim 25, characterized in that, Before the energy scheduling model sequentially predicts the prediction working parameters and target data corresponding to each sub-time period in the target time period to obtain the sub-strategy corresponding to each sub-time period, the method further includes: The second simulation model in the simulation model performs simulation and deduction on the prediction working parameters corresponding to the previous sub-period and the power prediction data in the target data to obtain the prediction working parameters corresponding to the sub-period, and sends the prediction working parameters corresponding to the sub-period to the energy dispatch model.

27. The energy dispatching method according to claim 25 or 26, characterized in that, The controller obtains the target scheduling strategy corresponding to the target application object based on the simulation results, including: Based on the simulation results, calculate the evaluation score corresponding to the simulation results from at least one dimension; The scheduling strategy corresponding to the simulation result with the highest evaluation score is determined as the target scheduling strategy.

28. An energy dispatching device, characterized in that, include: The first processing module is used to obtain the current sub-period and the prediction working parameters and target data corresponding to each remaining sub-period in the target time period; The second processing module is used to sequentially input the prediction working parameters and target data corresponding to each sub-time period into the energy scheduling model to obtain the sub-strategy corresponding to each sub-time period. The third processing module is used to sequentially input the sub-strategies corresponding to the sub-time periods into the first simulation model to obtain the sub-execution results corresponding to each sub-time period. The fourth processing module is used to adjust the model parameters of the energy scheduling model based on the simulation results obtained from the sub-execution results corresponding to each sub-time period in the target time period, and to re-execute the step of obtaining the prediction working parameters and target data corresponding to the current sub-time period and the remaining sub-time periods in the target time period until the preset stopping condition is reached. The fifth processing module is used to filter and obtain the target scheduling strategy corresponding to the target application object based on the simulation results.

29. An energy dispatching device, characterized in that, The device is applied to a simulation model, the simulation model including a first simulation model; the device includes: The sixth processing module is used to, upon receiving the sub-strategy corresponding to the sub-period in the target time period sent by the energy scheduling model, enable the first simulation model to perform simulation deduction based on the sub-strategy to obtain the sub-execution result corresponding to the sub-period; wherein, the sub-execution result is used to determine the simulation result corresponding to the target time period, the target time period includes multiple sub-periods, and the simulation result is used to adjust the model parameters of the energy scheduling model to obtain the target scheduling strategy corresponding to the target application object; The sub-strategy corresponding to the sub-period is predicted by the energy scheduling model based on the prediction working parameters and target data corresponding to the sub-period.

30. An energy dispatching device, characterized in that, The device, applied to energy dispatch models, includes: The seventh processing module is used to enable the energy scheduling model to predict the sub-strategy corresponding to each sub-time period based on the prediction working parameters corresponding to each sub-time period and the target data when it receives the prediction working parameters corresponding to each sub-time period in the target time period; the target time period includes multiple sub-time periods; The eighth processing module is used to enable the energy scheduling model to send the sub-strategies corresponding to the sub-time periods to the first simulation model in sequence, so that the first simulation model can execute the sub-strategies, obtain the sub-execution results corresponding to each sub-time period, and obtain the simulation results corresponding to the target time period based on each sub-execution result; The ninth processing module is used to, after obtaining the simulation results corresponding to the target time period and adjusting the model parameters of the energy scheduling model, cause the energy scheduling model to re-execute the step of predicting the sub-strategies corresponding to each sub-time period based on the prediction working parameters corresponding to each sub-time period and the target data, until a preset stopping condition is reached; so as to filter the target scheduling strategy corresponding to the target application object according to the obtained simulation results.

31. An energy dispatching device, characterized in that, An apparatus for use in an energy dispatching system, the energy dispatching system comprising a controller, an energy dispatching model, and a simulation model deployed on a simulation service cluster, wherein the simulation model and the energy dispatching model are connected; the apparatus comprises: The tenth processing module is used to enable the energy scheduling model to sequentially predict the prediction working parameters and target data corresponding to each sub-period in the target time period, so as to obtain the sub-strategy corresponding to each sub-period; the target time period includes multiple sub-periods; The eleventh processing module is used to enable the first simulation model in the simulation model to perform simulation deduction on each of the sub-strategies in sequence, and to obtain the sub-execution results corresponding to each of the sub-time periods; The twelfth processing module is used to enable the controller to adjust the model parameters of the energy scheduling model when the simulation results are obtained based on the sub-execution results corresponding to each sub-time period in the target time period. The thirteenth processing module is used to enable the adjusted energy scheduling model to re-execute the steps of predicting the prediction working parameters and target data corresponding to each sub-period in the target time period in turn, and obtaining the sub-strategy corresponding to each sub-period, until the preset stopping condition is reached. The fourteenth processing module is used to enable the controller to filter and obtain the target scheduling strategy corresponding to the target application object based on the simulation results.

32. An energy dispatching system, characterized in that, include: Controller; A simulation model is deployed on a simulation service cluster. The controller is connected to the simulation service cluster and controls the target computing node in the simulation service cluster to enable the model in the simulation model corresponding to the target computing node to perform simulation and deduction. An energy scheduling model is connected to both the simulation model and the controller. The controller is configured to execute the energy scheduling method as described in any one of claims 1-15.

33. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the energy scheduling method as described in any one of claims 1-27.

34. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the energy scheduling method as described in any one of claims 1-27.

35. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the energy scheduling method as described in any one of claims 1-27.

Images