Simulation verification method and system for optical storage direct-flexible micro-grid and computer program product

Abstract

The invention discloses a simulation verification method and system for an optical storage direct-flexible micro-grid and a computer program product, and relates to the technical field of scheduling optimization algorithms for power and energy management. The simulation verification method for the optical storage direct-flexible micro-grid comprises the following steps: acquiring historical data; a scheduling instruction and plan data are generated according to the historical data, and the plan data is an assumed effect value obtained by executing the scheduling instruction within a predicted set duration. And according to the historical data, generating a simulation condition of a set duration, wherein the simulation condition comprises a first prediction environment parameter, first prediction electricity price data and first prediction load data. And executing the scheduling instruction according to the simulation condition and generating simulation data. And comparing the planning data with the simulation data to evaluate the performance of the scheduling instruction. By adopting the technology provided by the invention, a diversified simulation environment can be effectively constructed to evaluate the scheduling algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of dispatch optimization algorithms for power and energy management, and in particular to a simulation verification method, system and computer program product for a photovoltaic-storage-direct-flexible microgrid. Background Art

[0002] Photovoltaic-storage-direct-flexible refers to a new type of power system that integrates photovoltaic power generation (photovoltaic), energy storage system (storage), direct power supply (direct) and flexible power consumption (flexible). It improves the stability of energy supply through the complementarity of photovoltaic power generation and energy storage system, and improves energy utilization based on the direct supply of electricity and the flexible regulation ability of power consumption equipment. Among them, the energy management system (EMS) can monitor, control and optimize the energy flow and information flow of the entire photovoltaic-storage-direct-flexible system.

[0003] For example, a Chinese invention patent document (CN118539483A) discloses a method for managing energy in a solar-storage, direct-flexible park. By controlling the equipment in each park, it can achieve energy scheduling in multiple parks with minimal network loss and the lowest energy equipment operating cost in a single park. However, after solving the control instructions based on the constraint rules, it directly sends them to the target park at the equipment layer, lacking the necessary verification means. Although it monitors and feedbacks operating data in real time for the optimization strategy, it cannot foresee and respond to future changes in a timely manner, resulting in a certain degree of blindness in the scheduling strategy.

[0004] Based on this, the rationality of the dispatching instructions can be determined by simple difference verification or by establishing a topological structure based on real historical data (such as weather, photovoltaic power generation information, historical electricity price information, and charging pile historical operation data over the past period of time). However, this verification method has problems such as single data and low environmental variables, resulting in low accuracy and robustness of the verification, which in turn makes it difficult for the dispatching instructions to adapt to the uncertainty and dynamic changes in the operation of the microgrid.

[0005] In addition, a method of building a system simulation model to input instructions for verification has been proposed. MATLAB / Simulink is used as a simulation platform, and the existing module library is used to create a system simulation model. The simulation results are analyzed to verify the feasibility of the scheduling instructions. However, this method requires a long simulation run time, high computing resource usage, and reliance on the existing module library. The simulation accuracy of the external environment is low, which also leads to low accuracy and robustness of verification. Summary of the invention

[0006] The present invention provides a simulation verification method, system and computer program product for a photovoltaic-storage-direct-flexible microgrid to solve the problems of single data and low environmental variables in the verification method in the prior art.

[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is to provide a simulation verification method of a photovoltaic storage direct flexible microgrid, and the simulation verification method of the photovoltaic storage direct flexible microgrid comprises: Acquire historical data, including historical photovoltaic data, historical electricity price data, and historical load data.

[0008] The scheduling instructions and plan data are generated according to the historical data, wherein the plan data is a hypothetical effect value obtained by executing the scheduling instructions within a predicted set time period.

[0009] A simulation condition of a set duration is generated according to the historical data, wherein the simulation condition includes a first predicted environmental parameter, a first predicted electricity price data, and a first predicted load data.

[0010] The scheduling instruction is executed according to the simulation condition and simulation data is generated.

[0011] The planned data is compared to the simulated data to evaluate the performance of the dispatch instruction.

[0012] The technical solution provided by the present invention has the following beneficial effects compared with the prior art: The dispatching instructions and simulation conditions with the same set duration are generated independently through historical data, and the dispatching instructions are executed according to the simulation conditions to simulate the actual power grid operation, and the results (i.e., simulation data) are used as the actual situation of the simulation, and then the performance of the dispatching instructions under various simulation conditions is measured; the economic effect of the dispatching instructions can be evaluated by comparing them with the results (i.e., planned data) expected when the dispatching instructions are generated. This allows staff to optimize the dispatching strategy before actual deployment and reduce actual operation risks.

[0013] Among them, compared with the current single data source method, the above-mentioned simulation conditions of a set duration are generated based on historical data, including simulating and generating the first predicted environmental parameters and the first predicted electricity price data to simulate the possible future environment, and generating the first predicted load data to simulate the conditions that may be encountered in the future. The combination can effectively build a more comprehensive, more diverse, and more realistic simulation environment, thereby improving the accuracy and reliability of the verification process. In addition, based on the comparison of simulation data and planned data, it is convenient for staff to determine the weak areas of the scheduling algorithm while evaluating the scheduling strategy, and then can make targeted optimization adjustments.

[0014] In some embodiments, generating simulation conditions of a set duration according to the historical data includes: The first photovoltaic prediction model, the first electricity price prediction model and the first load prediction model are pre-built; the first photovoltaic prediction model generates the first predicted photovoltaic data of the set duration according to the historical photovoltaic data; the first electricity price prediction model generates the first electricity price prediction data of the set duration according to the historical electricity price data; the first load prediction model generates the first load prediction data of the set duration according to the historical load data. To generate future environmental conditions that cover as many possible conditions as possible.

[0015] In some embodiments, generating scheduling instructions and planning data based on the historical data includes: Generate second predicted photovoltaic data, second predicted electricity price data and second predicted load data for a set time period according to the historical data; generate the dispatch instruction according to the second predicted photovoltaic data, the second predicted electricity price data and the second predicted load data using the EMS dispatch algorithm; calculate and execute the dispatch instruction to generate plan data under the conditions of the second predicted photovoltaic data, the second predicted electricity price data and the second predicted load data within the set time period.

[0016] In some embodiments, executing the scheduling instruction according to the simulation condition and generating simulation data further includes: A pre-constructed microgrid model is a mathematical model constructed based on photovoltaic physical parameters, energy storage physical parameters and load physical parameters; the microgrid model calculates photovoltaic power generation simulation data, charge and discharge simulation data and charging pile operation data of the set duration according to the scheduling instructions and the simulation conditions; and the target completion rate is calculated based on the photovoltaic data, the charge and discharge simulation data and the charging pile operation simulation data.

[0017] In some embodiments, comparing the planned data with the simulation data to evaluate the performance of the scheduling instruction further includes: The set time length is adjusted in sequence, and multiple time scales of hourly, daily and monthly levels are generated; the prediction accuracy of the dispatching strategy model and the prediction accuracy of the microgrid model are adjusted respectively according to the multiple time scales to generate multiple planned data of hourly, daily and monthly levels and multiple simulated data of hourly, daily and monthly levels; the corresponding planned data and simulated data are compared in sequence to evaluate their prediction accuracy at different time scales.

[0018] In some embodiments, obtaining historical data includes: Acquire historical weather data and historical photovoltaic power generation data as historical photovoltaic data, wherein the historical weather data includes temperature, humidity, light intensity, light duration and cloud coverage; acquire historical electricity price data; acquire historical load data, wherein the historical load data includes historical charging pile operation data; pre-process the historical photovoltaic data, historical electricity price data and historical load data respectively, specifically: detect missing values ​​and fill the missing values ​​by difference or mean; detect outliers and remove outlier data points.

[0019] In some embodiments, comparing the planned data with the simulated data to evaluate the performance of the dispatch instruction includes: Calculate the power purchased and sold by the power grid , recorded as:

[0020] in, is the grid power, is the photovoltaic power generation power, is the energy storage charging and discharging power, is the load power; Calculate the revenue from buying and selling electricity from the power grid , recorded as:

[0021] in, is the grid power, is the predicted electricity price at time i, n represents the set recorded data point, and m represents the number of records within 1 hour; Calculate Load Benefit , recorded as:

[0022] in, Indicates the charging price; Calculate total revenue , recorded as:

[0023] Generate multiple sets of planned total revenue data according to the planned data ; Generate multiple sets of simulated total revenue data based on the simulated data Compare the corresponding total benefit data of the plan in turn and simulated total return data .

[0024] In some embodiments, the simulation verification method of the photovoltaic storage direct flexible microgrid further includes: A first scheduling strategy model is pre-constructed according to a first EMS algorithm, and a second scheduling strategy model is pre-constructed according to a second EMS algorithm; the first scheduling strategy model generates a first scheduling instruction and a first plan data according to the historical data, wherein the first plan data is an assumed effect value predicted by the first scheduling strategy model when executing the first scheduling instruction within the set time period.

[0025] The second scheduling strategy model generates a second scheduling instruction and a second planning data according to the historical data, wherein the second planning data is an assumed effect value predicted by the second scheduling strategy model when executing the second scheduling instruction within the set time period.

[0026] The first scheduling instruction and the second scheduling instruction are respectively executed according to the simulation conditions, and first simulation data and second simulation data are generated; the first simulation data and the second simulation data are compared to evaluate the performance of the first EMS scheduling algorithm and the second EMS scheduling algorithm.

[0027] In some embodiments, the present application also provides a simulation verification system for a photovoltaic-storage-direct-flexible microgrid, which applies the above-mentioned simulation verification method for a photovoltaic-storage-direct-flexible microgrid, including: a simulation system, a cloud platform, and an evaluation and measurement system. The simulation system includes a first prediction model and a microgrid model, wherein the first prediction model can generate simulation conditions of a set duration based on historical data predictions; the cloud platform and the microgrid model interact with each other through an API interface, wherein the cloud platform is configured with a scheduling strategy model, and the scheduling strategy model can generate scheduling instructions and plan data based on the historical data; the evaluation and measurement system can compare the plan data with the simulation data to evaluate the performance of the scheduling instructions.

[0028] By adopting the above technical solution and exchanging data through the API interface, the data source relied on by the scheduling strategy model is transmitted by the simulation system or read in real time from the cloud platform. Therefore, for the scheduling strategy model, each calculation is the same as the actual operation. This can verify the real performance of the cloud platform and reduce the error conditions of its actual deployment and operation.

[0029] In some embodiments, the present application also provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the simulation verification method of the above-mentioned photovoltaic-storage-direct-flexible microgrid. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work, among which: Figure 1 This is a flow chart of an embodiment of a simulation verification method for a photovoltaic storage direct-flexible microgrid provided by the present invention. Figure 1 ; Figure 2 This is a flow chart of an embodiment of a simulation verification method for a photovoltaic-storage-direct-flexible microgrid provided by the present invention. Figure 2 ; Figure 3 This is a flow chart of an embodiment of a simulation verification method for a photovoltaic-storage-direct-flexible microgrid provided by the present invention. Figure 3 ; Figure 4 This is a flow chart of an embodiment of a simulation verification method for a photovoltaic storage direct-flexible microgrid provided by the present invention. Figure 4 . DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0032] See also Figure 1 to Figure 2 As shown, Figure 1 The schematic diagram shows a process of an embodiment of a simulation verification method of a photovoltaic storage direct flexible microgrid provided by the present application. Figure 1 ; Figure 2 The schematic diagram shows a process of an embodiment of a simulation verification method of a photovoltaic storage direct flexible microgrid provided by the present application. Figure 2 .

[0033] In some embodiments, the simulation verification method of the photovoltaic storage direct flexible microgrid includes: Step S10, obtaining historical data, the historical data including historical photovoltaic data, historical electricity price data and historical load data.

[0034] Exemplarily, historical data includes historical weather data and historical photovoltaic power generation data. Historical weather data can be obtained through photovoltaic power generation detection equipment, meteorological stations and satellite data, including temperature, humidity, light intensity, light duration, cloud coverage, wind speed, etc., which are not limited in this application. Among them, historical electricity price data includes electricity prices in different time periods, such as peak and valley electricity prices, etc., which are obtained by power grid companies or electricity price databases; historical load data is the power load in a specific area, such as charging pile power and charging pile operation data, etc., which are obtained by real-time detection and recording of the load monitoring system.

[0035] Step S20, generating scheduling instructions and planning data based on historical data, wherein the planning data is an assumed effect value obtained by executing the scheduling instructions within a predicted set time period.

[0036] The existing EMS (energy management system) scheduling algorithm can be used to generate scheduling instructions and plan data based on historical data, wherein the scheduling instructions include power control, charging and discharging plan and load management of photovoltaic power generation. For example, after the scheduling instructions are output with rules as constraints of the EMS scheduling algorithm, it is predicted that the scheduling instructions can be executed within the set time to generate plan data.

[0037] Step S30, generating simulation conditions of a set duration according to historical data, the simulation conditions including first predicted environmental parameters, first predicted electricity price data and first predicted load data.

[0038] Use historical data to train prediction models (such as the first photovoltaic prediction model, the first electricity price prediction model, and the first load prediction model), and use the trained prediction model to generate the first predicted environmental parameters, the first predicted electricity price data, and the first predicted load data within a set time period in the future. Exemplarily, the construction of the prediction model can use AI prediction algorithms (such as LSTM, deep neural network, random forest, GBDT, gray prediction model, etc.). Among them, different prediction models can be used to generate predictions for simulation conditions. For example, the first photovoltaic prediction model can use LSTM, the first electricity price prediction model can use GBDT, and the first load prediction model can use the first time series analysis, according to different accuracy requirements and computing power selection, which is not limited in this application.

[0039] Step S40, executing the scheduling instruction according to the simulation conditions and generating simulation data.

[0040] The first predicted environmental parameters, first predicted electricity price data, and first predicted load data generated through independent prediction of historical data are set as the conditions of the simulation environment, and the scheduling instructions are executed in the simulation environment to simulate the actual operation of the photovoltaic storage direct flexible microgrid, and the results are used as simulation data to reflect the operating status of the photovoltaic storage direct flexible microgrid. For example, the generated scheduling instructions include a charging pile discharge adjustment plan, and the existing charging piles are measured under the changes of the first predicted electricity price, the first predicted environmental parameters, and the first predicted load data, and the results generated by the measurement are simulation data.

[0041] Step S50, comparing the planned data with the simulation data to evaluate the performance of the scheduling instruction.

[0042] When the difference between the planned data and the simulated data is small, it is generally considered that the data predicted by the scheduling algorithm is highly accurate. Conversely, when the difference between the planned data and the simulated data is large, it is considered that the data predicted by the scheduling algorithm is low accurate.

[0043] In the embodiment of the present application, the scheduling instructions and simulation conditions with the same set duration are generated through historical data prediction respectively to ensure the independence between the two, reduce data deviation and allow the scheduling instructions to be tested under simulation conditions as close as possible to the actual operating conditions, so as to enhance the robustness of the scheduling strategy. The results (i.e., simulation data) are used as the actual situation of the simulation, and then the performance of the scheduling instructions under various simulation conditions is measured; comparing it with the results (i.e., planned data) expected when the scheduling instructions are generated can evaluate the economic effect of the scheduling instructions. This allows staff to optimize the scheduling strategy before actual deployment to improve the accuracy of the generation of scheduling instructions and reduce actual operating risks.

[0044] By simulating and generating the first predicted environmental parameters and the first predicted electricity price data to simulate the possible future environment, the first predicted load data is generated to simulate the conditions that may be encountered in the future. The combination can effectively build a more comprehensive, more diverse and more realistic simulation environment, thereby improving the accuracy and reliability of the verification process.

[0045] In some embodiments, step S30, generating simulation conditions of a set duration according to historical data, includes: The first photovoltaic prediction model, the first electricity price prediction model and the first load prediction model are pre-constructed; the first photovoltaic prediction model generates the first predicted photovoltaic data of a set duration based on the historical photovoltaic data; the first electricity price prediction model generates the first electricity price prediction data of a set duration based on the historical electricity price data; the first load prediction model generates the first load prediction data of a set duration based on the historical load data.

[0046] In the embodiment of the present application, in some application scenarios, the photovoltaic data interval is conventionally measured once every 15 minutes. The first predicted photovoltaic data predicted by the first photovoltaic model is the average photovoltaic power generation power of photovoltaic power generation. The average power multiplied by the set time length is the predicted photovoltaic power generation.

[0047] Combination Figure 3 As shown, Figure 3 The schematic diagram shows a process of an embodiment of a simulation verification method of a photovoltaic storage direct flexible microgrid provided by the present application. Figure 3 .

[0048] In some embodiments, step S20, generating scheduling instructions and planning data based on historical data, includes: Step S200, generating second predicted photovoltaic data, second predicted electricity price data and second predicted load data of a set duration according to historical data.

[0049] The second photovoltaic prediction model, the second electricity price prediction model and the second load prediction model are pre-built, and are trained using historical data to generate second predicted photovoltaic data, second predicted electricity price data and second predicted load data. The second photovoltaic prediction model, the second electricity price prediction model and the second load prediction model are trained with the first photovoltaic prediction model, the first electricity price prediction model and the first load prediction model using realistic historical data to improve the authenticity and utilization of the data.

[0050] Step S210: Generate a dispatch instruction using the EMS dispatch algorithm according to the second predicted photovoltaic data, the second predicted electricity price data and the second predicted load data.

[0051] The goal of the EMS scheduling algorithm is to reasonably arrange the energy storage charging and discharging cycles and power output. It can set algorithm parameters (such as the objective function of minimizing cost) and constraints (such as supply and demand balance) to adjust load power consumption, etc.

[0052] Step S220, calculating the scheduling instruction to generate plan data within the set time period under the conditions of the second predicted photovoltaic data, the second predicted electricity price data and the second predicted load data.

[0053] Exemplarily, the dispatching instruction includes a photovoltaic dispatching instruction, an energy storage dispatching instruction, and a load dispatching instruction. In some application scenarios, the interval of the photovoltaic dispatching instruction is 15 minutes, and the energy storage dispatching instruction and the load dispatching instruction follow the photovoltaic dispatching instruction, that is, a dispatching value is generated every 15 minutes. In order to balance the load power consumption and photovoltaic power generation, the difference in electricity is the purchase or sale of the power grid. By calculating the electricity fee, the planned economic benefits of the dispatching instruction and indicators such as the self-generation and self-use rate can be calculated.

[0054] By constructing a simulation environment for evaluation in step S50, the execution effect can be accurately evaluated before the scheduling algorithm is configured, and the scheduling algorithm can be optimized according to the evaluation effect to continuously improve its accuracy and scheduling effect.

[0055] In some embodiments, step S40, executing the scheduling instruction and generating simulation data according to the simulation conditions, further includes: A pre-constructed microgrid model is a mathematical model constructed based on photovoltaic physical parameters, energy storage physical parameters and load physical parameters; the microgrid model calculates photovoltaic power generation simulation data, charging and discharging simulation data and charging pile operation data of a set duration according to the dispatching instructions and simulation conditions; and the target completion rate is calculated based on the photovoltaic data, charging and discharging simulation data and charging pile operation simulation data.

[0056] In the embodiment of the present application, the microgrid model simulates the components in the actual microgrid, including the photovoltaic system, energy storage system and load system, as well as the relationship between the components. For example, the photovoltaic physical parameters may include the photovoltaic panel area, power generation efficiency, etc.; the energy storage physical parameters include the battery capacity, charging and discharging efficiency, etc.; the load physical parameters include the number of charging piles, charging characteristics, etc.

[0057] Among them, the simulation conditions and dispatching instructions are input into the microgrid system to simulate the actual operation of the photovoltaic storage direct-flexible microgrid, and the operation results are simulation data.

[0058] In some embodiments, step S50, comparing the planned data with the simulated data to evaluate the performance of the dispatch instruction, further comprises: The set duration is adjusted in sequence, and multiple time scales at the hourly, daily and monthly levels are generated; the prediction accuracy of the dispatching strategy model and the prediction accuracy of the microgrid model are adjusted according to the multiple time scales to generate multiple hourly, daily and monthly plan data and multiple hourly, daily and monthly simulation data; the corresponding plan data and simulation data are compared in sequence.

[0059] In the embodiment of the present application, the prediction accuracy of the scheduling strategy model, the first prediction model and the microgrid model are adjusted on different time scales to simulate the scheduling results of the EMS scheduling algorithm from high to low prediction accuracy.

[0060] In some implementation schemes, step S10, obtaining historical data, includes preprocessing historical photovoltaic data, historical electricity price data and historical load data respectively, specifically: detecting missing values ​​and filling the missing values ​​by difference or mean; detecting outliers and removing abnormal data points.

[0061] In an embodiment of the present application, the above data are preprocessed to ensure the integrity and accuracy of the data, thereby improving the prediction accuracy of the first prediction model and the scheduling strategy model. Exemplarily, the data preprocessing also includes data normalization and feature vector extraction.

[0062] In some embodiments, step S50, comparing the planned data with the simulated data to evaluate the performance of the dispatch instruction, includes: Calculate the power purchased and sold by the power grid , recorded as:

[0063] in, is the grid power, is the photovoltaic power generation power, is the energy storage charging and discharging power. It is negative when it is on and positive when it is discharged; is the load power. For example, the load power of a charging pile without V2G (Vehicle-to-Grid) function is defined as a positive value.

[0064] Calculate the revenue from buying and selling electricity from the power grid , recorded as:

[0065] in, is the grid power, is the predicted electricity price at time i, n represents the set recorded data point, and m represents the number of records within 1 hour. For example, if the calculation is performed every 15 minutes, there are 96 record points in a day, then .

[0066] Calculate Load Benefit , recorded as:

[0067] in, Indicates the charging price; Calculate total revenue , recorded as:

[0068] Generate multiple sets of planned total benefit data based on planned data ; Generate multiple sets of simulated total return data based on simulated data Compare the corresponding total benefit data of the plan in turn and simulated total return data .

[0069] In the embodiment of the present application, the accuracy of the EMS scheduling algorithm is determined by respectively calculating the planned total revenue data and the simulated total revenue data and comparing the difference between the two.

[0070] Combination Figure 4 As shown, Figure 4 The schematic diagram shows a process of an embodiment of a simulation verification method of a photovoltaic storage direct flexible microgrid provided by the present application. Figure 4 .

[0071] In some embodiments, the simulation verification method of the photovoltaic storage direct flexible microgrid further includes: Pre-constructing a first scheduling strategy model according to a first EMS algorithm, and pre-constructing a second scheduling strategy model according to a second EMS algorithm; Step S21, the first scheduling strategy model generates a first scheduling instruction and first plan data based on historical data, wherein the first plan data is an assumed effect value obtained by executing the first scheduling instruction within a set time period predicted by the first scheduling strategy model.

[0072] Step S22, the second scheduling strategy model generates a second scheduling instruction and second plan data based on the historical data, wherein the second plan data is an assumed effect value obtained by executing the second scheduling instruction within a set time period predicted by the second scheduling strategy model.

[0073] Step S41, executing the first scheduling instruction and the second scheduling instruction respectively according to the simulation conditions, and generating the first simulation data and the second simulation data; Step S51 : comparing the first simulation data and the second simulation data to evaluate the performance of the first EMS scheduling algorithm and the second EMS scheduling algorithm.

[0074] In the embodiment of the present application, the formation gap between EMS algorithm 1 and EMS algorithm 2 can also be compared through a simulation environment, and the first scheduling instruction and the second scheduling instruction respectively generated by them can be run under the same simulation conditions to evaluate the performance difference between the two.

[0075] Table 1: Profit analysis of EMS algorithm 1 and EMS algorithm 2

[0076] For example, as shown in Table 1, EMS algorithm 1 and EMS algorithm 2 repeatedly generate 10 sets of simulation environments and dispatch instructions in the microgrid model and calculate the corresponding total benefits in turn as above. By comparison, it can be obtained that the economic benefit performance of the first dispatch instruction generated by EMS algorithm 1 is greater than the economic benefit performance of the second dispatch instruction generated by EMS algorithm 2.

[0077] In some embodiments, the present application further provides a simulation verification system for a photovoltaic-storage-direct-flexible microgrid, which applies the simulation verification method for the photovoltaic-storage-direct-flexible microgrid, including: A simulation system, the simulation system includes a first prediction model and a microgrid model, the first prediction model can generate simulation conditions of a set duration based on historical data prediction; Cloud platform: The cloud platform and the microgrid model interact with each other through the API interface. The cloud platform is equipped with a dispatch strategy model, which can generate dispatch instructions and plan data based on historical data. Evaluation and measurement system, the evaluation and measurement system can compare the planned data with the simulation data to evaluate the performance of the scheduling instructions.

[0078] In the embodiment of the present application, data interaction is performed through the API interface, so that the data source relied on by the scheduling strategy model is transmitted by the simulation system or read in real time from the cloud platform, so that for the scheduling strategy model, each measurement performed is the same as the actual operation. It can also verify the real performance of the cloud platform and reduce the error conditions of its actual deployment and operation.

[0079] In some embodiments, the present application also provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the simulation verification method of the above-mentioned photovoltaic-storage-direct-flexible microgrid.

[0080] It should be understood by those skilled in the art that the embodiments of the present invention can be provided as methods, systems or computer program products. Those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.

[0081] The above description is only an implementation mode of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly used in other related technical fields, should be included in the protection scope of the present invention.

Claims

1. A simulation verification method for a photovoltaic-storage direct-flexible microgrid, characterized in that: include: Acquiring historical data, wherein the historical data includes historical photovoltaic data, historical electricity price data, and historical load data; Generate scheduling instructions and planning data according to the historical data, wherein the planning data is a hypothetical effect value obtained by executing the scheduling instructions within a predicted set time period; Generate a simulation condition of a set duration according to the historical data, the simulation condition including a first predicted environmental parameter, a first predicted electricity price data and a first predicted load data; Execute the scheduling instruction according to the simulation condition and generate simulation data; The planned data is compared to the simulated data to evaluate the performance of the dispatch instruction.

2. The simulation verification method of the photovoltaic storage direct flexible microgrid according to claim 1 is characterized in that: The generating of simulation conditions of a set duration according to the historical data comprises: Pre-building a first photovoltaic prediction model, a first electricity price prediction model, and a first load prediction model; The first photovoltaic prediction model generates first predicted photovoltaic data of the set duration according to the historical photovoltaic data; The first electricity price prediction model generates first electricity price prediction data of the set duration according to the historical electricity price data; The first load forecasting model generates the first load forecasting data of the set duration according to the historical load data.

3. The simulation verification method of the photovoltaic storage direct flexible microgrid according to claim 1 is characterized in that: The generating of the scheduling instructions and planning data according to the historical data comprises: Generate second predicted photovoltaic data, second predicted electricity price data and second predicted load data of a set duration according to the historical data; Generate the dispatch instruction according to the second predicted photovoltaic data, the second predicted electricity price data and the second predicted load data using an EMS dispatch algorithm; Calculate the scheduling instruction to generate plan data within the set time period under the conditions of the second predicted photovoltaic data, the second predicted electricity price data and the second predicted load data.

4. The simulation verification method of the photovoltaic energy storage direct flexible microgrid according to claim 1 is characterized in that: The step of executing the scheduling instruction according to the simulation condition and generating simulation data further includes: Pre-constructing a microgrid model, wherein the microgrid model is a mathematical model constructed based on photovoltaic physical parameters, energy storage physical parameters and load physical parameters; The microgrid model calculates the photovoltaic power generation simulation data, the charging and discharging simulation data and the charging pile operation data of the set duration according to the dispatching instruction and the simulation conditions; The target completion rate is calculated based on the photovoltaic data, the charging and discharging simulation data and the charging pile operation simulation data.

5. The simulation verification method of the photovoltaic energy storage direct flexible microgrid according to claim 4 is characterized in that: The comparing the planned data with the simulation data to evaluate the performance of the dispatch instruction further includes: The set duration is adjusted in sequence, and a plurality of time scales at the hour level, the day level and the month level are generated; According to the multiple time scales, the prediction accuracy of the dispatching strategy model and the prediction accuracy of the microgrid model are adjusted respectively to generate the planned data at multiple hourly, daily and monthly levels and the simulation data at multiple hourly, daily and monthly levels; The corresponding planned data and simulation data are compared in sequence.

6. The simulation verification method of the photovoltaic storage direct flexible microgrid according to claim 1 is characterized in that: The obtaining of historical data includes: Acquire historical weather data and historical photovoltaic power generation data as historical photovoltaic data, wherein the historical weather data includes temperature, humidity, light intensity, light duration, and cloud coverage; Get historical electricity price data; Acquire historical load data, wherein the historical load data includes historical charging pile operation data; The historical photovoltaic data, historical electricity price data and historical load data are preprocessed respectively, specifically: Detect missing values ​​and fill in the missing values ​​by difference or mean; Detect outliers and remove anomalous data points.

7. The simulation verification method of the photovoltaic energy storage direct flexible microgrid according to claim 1 is characterized in that: The comparing the planned data with the simulation data to evaluate the performance of the dispatch instruction comprises: Calculate the power purchased and sold by the power grid , recorded as: in, is the grid power, is the photovoltaic power generation power, is the energy storage charging and discharging power, is the load power; Calculate the revenue from buying and selling electricity from the power grid , recorded as: in, is the grid power, is the predicted electricity price at time i, n represents the set recorded data point, and m represents the number of records within 1 hour; Calculate Load Benefit , recorded as: in, Indicates the charging price; Calculate total revenue , recorded as: Generate multiple sets of planned total revenue data according to the planned data ; Generate multiple sets of simulated total revenue data based on the simulated data Compare the corresponding total benefit data of the plan in turn and simulated total return data .

8. The simulation verification method of the photovoltaic-storage-direct-flexible microgrid according to any one of claims 1 to 7, characterized in that: The simulation verification method of the photovoltaic storage direct flexible microgrid also includes: Pre-constructing a first scheduling strategy model according to a first EMS algorithm, and pre-constructing a second scheduling strategy model according to a second EMS algorithm; The first scheduling strategy model generates a first scheduling instruction and first plan data according to the historical data, wherein the first plan data is an assumed effect value obtained by executing the first scheduling instruction within the set time period predicted by the first scheduling strategy model; The second scheduling strategy model generates a second scheduling instruction and a second planning data according to the historical data, wherein the second planning data is an assumed effect value obtained by executing the second scheduling instruction within the set time period predicted by the second scheduling strategy model; Execute the first scheduling instruction and the second scheduling instruction respectively according to the simulation condition, and generate first simulation data and second simulation data; The first simulation data and the second simulation data are compared to evaluate performance of the first EMS scheduling algorithm and the second EMS scheduling algorithm.

9. A simulation verification system for a photovoltaic-storage-direct-flexible microgrid, characterized in that: The simulation verification method of the photovoltaic storage direct flexible microgrid according to any one of claims 1 to 8 comprises: A simulation system, the simulation system comprising a first prediction model and a microgrid model, the first prediction model being capable of generating simulation conditions of a set duration based on historical data prediction; A cloud platform, wherein the cloud platform and the microgrid model perform data interaction through an API interface, wherein the cloud platform is configured with a scheduling strategy model, and the scheduling strategy model can generate scheduling instructions and plan data according to the historical data; An evaluation and calculation system is capable of comparing the plan data with the simulation data to evaluate the performance of the scheduling instruction.

10. A computer program product, characterized in that It includes computer programs / instructions, which, when executed by a processor, implement the simulation verification method of the photovoltaic-storage-direct-flexible microgrid as described in any one of claims 1 to 8.

Images