Optimized scheduling method of photovoltaic and commercial power combined power supply system
By optimizing the scheduling of the photovoltaic and grid-connected power supply system through data processing and model prediction, the problem of balancing power supply stability and economy is solved, and the energy consumption of the distributed computing system is optimized, maximizing the utilization of photovoltaic power and reducing grid power consumption.
Patent Information
- Application Number
- CN202511710433.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-20
- Publication Date
- 2026-02-27
AI Technical Summary
Existing technologies are insufficient to effectively optimize the scheduling of photovoltaic and grid-connected power supply systems, making it difficult to balance power supply stability and economy, and the energy consumption problem of distributed computing systems has not been effectively solved.
By collecting and processing historical power generation and meteorological data, a long short-term memory network model is used to predict photovoltaic power. Combined with the relationship between the power consumption and computing speed of the computer group, the scheduling strategy is optimized to determine the frequency regulation time period and computing time. The overall power supply total cost objective function is constructed to solve for the optimal solution.
It achieves the goal of maximizing the utilization of photovoltaic power, reducing grid power consumption costs, and optimizing energy efficiency while meeting the requirements of computing tasks and ensuring power supply stability.
Smart Images
Figure CN121584759A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of energy system optimization and scheduling technology, and in particular to an optimization and scheduling method for a photovoltaic and municipal power combined power supply system. Background Technology
[0002] New energy power supply, such as photovoltaic (PV) and wind power, is affected by temperature, weather, and sunlight, resulting in strong randomness and volatility in power supply. However, under the "dual carbon" (carbon dioxide, carbon emissions, and carbon sequestration) context, new energy sources are renewable energy sources, and their power generation process is clean and environmentally friendly, helping to reduce air pollution and alleviate environmental problems. Furthermore, in recent years, the cost of new energy power generation, especially PV power generation, has gradually decreased, demonstrating good economic viability. Combining grid power with new energy power generation can improve power supply stability and reduce costs, making it a choice that meets the demands of the times. However, this combination requires precise and comprehensive optimized dispatch strategies to maximize PV absorption, reduce curtailment rates, and simultaneously ensure the stability and economy of the power supply system.
[0003] Meanwhile, with the rapid development of information technology, distributed computing has become a key technology for processing large-scale data and complex computing tasks. Distributed computing plays an important role in modern information technology, especially in processing large-scale data and complex computing tasks. However, the widespread application of such systems brings significant energy consumption challenges. In distributed computing environments, energy demands become diverse and complex due to the wide distribution of nodes and dynamic load changes. The continuous operation of energy-intensive equipment and uneven load distribution further exacerbate the energy consumption problem. Optimizing the integrated energy system of distributed computing aims to improve system energy efficiency and reduce operating costs through intelligent scheduling and resource management.
[0004] In this context, optimized scheduling strategies for energy systems are particularly important. Intelligent scheduling algorithms can dynamically adjust the ratio of photovoltaic (PV) power to grid power under different load conditions, thereby maximizing PV energy utilization and reducing overall operating costs. Furthermore, considering the time sensitivity and priority of computational tasks, scheduling strategies must also be flexible and adaptable to ensure optimal system performance.
[0005] Therefore, it is necessary to provide a new technical solution to improve one or more of the problems existing in the above solutions.
[0006] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention
[0007] The purpose of this application is to provide an optimized scheduling method for a photovoltaic and grid-connected power supply system, thereby overcoming, to at least some extent, one or more problems caused by the limitations and defects of related technologies.
[0008] An optimized scheduling method for a photovoltaic and grid-connected power supply system according to an embodiment of this application includes: Collect historical power generation data, historical meteorological data, computer computing power related data, and real-time meteorological data of photovoltaic power plants; the computer computing power related data includes the computer frequency modulation range, the power consumption value corresponding to each frequency within the frequency modulation range, and the calculation speed corresponding to each power consumption value; Data processing is performed on the historical meteorological data and the historical power generation data to obtain a dataset; The dataset is divided into a training set, a validation set, and a test set. The long short-term memory network model is trained using the training set, the trained long short-term memory network model is validated using the validation set, and the validated long short-term memory network model is evaluated using the test set to obtain the photovoltaic power prediction model. The real-time meteorological data is input into the photovoltaic power prediction model to output the photovoltaic power prediction values at each time point in the future period, and then fitted into a photovoltaic daily power curve. A relationship curve between frequency and power consumption value is plotted based on the power consumption value corresponding to each frequency within the frequency modulation range, and is denoted as the first relationship curve. A relationship curve between frequency and calculation speed is plotted based on the power consumption value corresponding to each frequency within the frequency modulation range and the calculation speed corresponding to each power consumption value, and is denoted as the second relationship curve. A set of compact power consumption values is obtained based on the first relationship curve and the second relationship curve. The set of compact power consumption values can cover the power consumption value range. Based on the compact power consumption value and the photovoltaic daily power curve, the frequency modulation period for which the computer group needs to be frequency-modulated is determined; Based on the first relationship curve, the second relationship curve, and the frequency modulation time period, calculate the calculation time corresponding to each power value in the group of compact power consumption values, and record the calculation time corresponding to each power value in the group of compact power consumption values as the actual calculation time; Based on the actual calculation time, mains power supply, and photovoltaic power supply, a comprehensive total power supply cost objective function is constructed, and the optimal solution of the comprehensive total power supply cost objective function is obtained.
[0009] In the embodiments of this application, the step of determining the frequency modulation period for which the computer group needs frequency modulation based on the group's compact power consumption value and the photovoltaic daily power curve includes: By combining the compact power consumption values and the photovoltaic daily power curve, the intersection expression of the two is obtained. Solving the intersection expression yields the frequency modulation start time. and the end time of frequency modulation The frequency modulation time period is ~ The intersection point expression is as follows: (1) In the formula, This indicates the compact power consumption value of the group. The expression representing the daily power curve of a photovoltaic system.
[0010] In the embodiments of this application, the step of calculating the calculation time corresponding to each power consumption value in the group of compact power consumption values based on the first relationship curve, the second relationship curve, and the frequency modulation time period, and recording the calculation time corresponding to each power consumption value in the group of compact power consumption values as the actual calculation time, includes: Based on the first relationship curve, the frequency corresponding to each value in this set of compact power consumption values is obtained; Based on the second relationship curve, when the frequency is adjusted to the maximum value, the maximum computing speed corresponding to the computer group is obtained, and the shortest computing time is determined according to the maximum computing speed.
[0011] In the embodiments of this application, the step of calculating the calculation time corresponding to each power consumption value in the group of compact power consumption values based on the first relationship curve, the second relationship curve, and the frequency modulation time period, and recording the calculation time corresponding to each power consumption value in the group of compact power consumption values as the actual calculation time, includes: When the computer group is running at maximum computing speed, the time point for completing the computing task is between 0 and 10. The actual calculation time is obtained based on the shortest calculation time; wherein, the expression for the actual calculation time is as follows: (2) In the formula, Indicates the actual calculation time. Indicates the shortest calculation time. This indicates that the initial power consumption of the computer group has dropped to [a certain value]. The increase in the shortest calculation time is recorded as the increase in time. Based on the maximum computing speed and the shortest computing time, the total relative workload of the computer group is obtained; wherein, the expression for the total relative workload of the computer group is as follows: (3) In the formula, This represents the total relative workload of the computer group. Indicates the maximum calculation speed; when At that time, based on the maximum computing speed and the frequency modulation time period, the assumed relative workload of the computer group within the frequency modulation time period is obtained; wherein, the expression for the assumed relative workload within the frequency modulation time period is as follows: (4) In the formula, This represents the hypothetical relative workload within the frequency modulation period, and n represents the number of full days that the computer group spans while performing the computing tasks. Based on the total relative workload of the computer group and the assumed relative workload during the frequency modulation period, the assumed relative workload of the computer group during the non-frequency modulation period is obtained; wherein, the expression for the assumed relative workload during the non-frequency modulation period is as follows: (5) In the formula, This represents the assumed relative workload during non-frequency modulation periods; Based on the assumed relative workload during the frequency modulation period, the maximum computing speed, the assumed relative workload during the non-frequency modulation period, and the computing speed during the non-frequency modulation period, the assumed computing time is obtained; wherein, the expression for the assumed computing time is as follows: (6) In the formula, This indicates the assumed calculation time. This indicates the calculation speed during non-frequency modulation periods; Determine whether the assumed calculation duration extends into the next frequency modulation time period; like Then the actual calculation time is equal to the assumed calculation time, that is ; like ,Right now Then it is assumed that the calculation time has been extended into the next frequency modulation time period; when If the time point for completing the calculation task falls within the frequency modulation time period, then the expression for the actual calculation duration is as follows; (7) In the formula, This indicates the actual relative workload within the frequency modulation period; when If the time point for completing the calculation task falls within the time period after the end of the frequency modulation period, then the expression for the actual calculation duration is as follows: (8).
[0012] In the embodiments of this application, the step of calculating the calculation time corresponding to each power consumption value in the group of compact power consumption values based on the first relationship curve, the second relationship curve, and the frequency modulation time period, and recording the calculation time corresponding to each power consumption value in the group of compact power consumption values as the actual calculation time, includes: When the computer group is running at maximum computing speed, the time point for completing the computing task is within the frequency modulation time period, and the expression for the actual computing time is the same as formula (2); The final computation time is obtained based on the shortest computation time; wherein, the final computation time is the computation time on the last day of the shortest computation time for the computer group to complete the computation task at the maximum computation speed, and the expression for the final computation time is as follows: , (9) In the formula, Indicates the final calculation duration; when Then, the expression for the assumed relative workload within the frequency modulation time period is as follows: (10) At this point, the expression for the assumed calculation time is as follows: (11) like ,Right now Then the actual calculation time is equal to the assumed calculation time, that is ; like ,Right now The expression for the actual relative workload within the frequency modulation time period is as follows: (12) The expression for the actual calculation time is as follows: (13).
[0013] In the embodiments of this application, the step of calculating the calculation time corresponding to each power consumption value in the group of compact power consumption values based on the first relationship curve, the second relationship curve, and the frequency modulation time period, and recording the calculation time corresponding to each power consumption value in the group of compact power consumption values as the actual calculation time, includes: When the computer group is running at maximum computing speed, and the time point for completing the computing task is located at At this time, the expression for the assumed relative workload during the frequency modulation period is the same as formula (4), the expression for the assumed relative workload during the non-frequency modulation period is the same as formula (5), and the expression for the assumed calculation duration is the same as formula (6). like ,Right now Then the actual calculation time is equal to the assumed calculation time, that is ; like ,Right now The expression for the actual relative workload within the frequency modulation time period is as follows: (14) The expression for the actual calculation time is as follows: (15) like At this time, the expression for the relative workload within the frequency modulation time period is the same as that of formula (12), and the expression for the actual calculation duration is the same as that of formula (13).
[0014] In the embodiments of this application, the daily mains power purchase amount of the integrated power supply system is obtained based on the daytime mains power purchase amount and the nighttime mains power purchase amount; wherein, the expression for the daily mains power purchase amount of the integrated power supply system is as follows: (16) In the formula, This represents the daily amount of mains electricity purchased by the integrated power supply system. This indicates the amount of electricity purchased during the day. This indicates the amount of electricity purchased at night. When the maximum power consumption value in this set of compact power consumption values is greater than or equal to the maximum daily photovoltaic power, the expression for the daytime grid electricity purchase quantity is as follows: (17) In the formula, This indicates the maximum power consumption value within this group of compact power consumption values. This indicates the maximum daily power output of photovoltaic power. The value is 6:00, which is recorded as 0 in integral calculations. It is 18:00, which is recorded as 13 in integral calculation; When the maximum power consumption value in this set of compact power consumption values is less than the maximum daily photovoltaic power, by Solve for the start time of photovoltaic power supply and the end time of photovoltaic power supply The period during which only the photovoltaic power supply is ~ ,exist ~ The expression for the daytime electricity purchase volume is as follows: (18) The expression for the nighttime electricity purchase volume is as follows: (19).
[0015] In the embodiments of this application, the photovoltaic power supply cost is calculated based on the photovoltaic power supply and the levelized cost of electricity of the photovoltaic power station; wherein, the expression for the daily photovoltaic power supply cost of the integrated power supply system is as follows: (20) In the formula, This indicates the cost per kilowatt-hour of a photovoltaic power plant. This represents the daily photovoltaic power supply of the integrated power supply system. This indicates the daily photovoltaic power supply cost of the integrated power supply system.
[0016] In the embodiments of this application, the expression for the comprehensive power supply cost over n full days is as follows: (twenty one) In the formula, This represents the total cost of electricity supply over n complete days. This represents the daily cost of mains electricity purchased by the integrated power supply system, where 's' is the unit price of electricity purchased from the main grid. This indicates the levelized cost of electricity (LCOE) of a photovoltaic system. The expression for the actual calculation time is as follows: (twenty two) When the maximum power consumption value in this set of compact power consumption values is greater than or equal to the maximum daily photovoltaic power, the expression for the power supply cost over the remaining time is as follows: (twenty three) When the maximum power consumption value in this set of compact power consumption values is less than the maximum daily photovoltaic power, the expression for the power supply cost during the remaining time is as follows: (twenty four) In the formula, This indicates the cost of supplying electricity for the remaining duration.
[0017] In the embodiments of this application, the expression of the overall power supply total cost objective function is as follows: (25) In the formula, This represents the objective function for the total cost of integrated power supply.
[0018] The technical solutions provided by the embodiments of this application may include the following beneficial effects: In one embodiment of this application, real-time meteorological data is input into a photovoltaic power prediction model using the above method to output photovoltaic power prediction values at various time points, which are then fitted into a photovoltaic daily power curve. Then, using a set of compact power consumption values and the photovoltaic daily power curve, the frequency regulation time period required for the computer group can be determined. The actual calculation duration is obtained based on the first relationship curve, the second relationship curve, and the frequency regulation time period. A comprehensive power supply total cost objective function is constructed based on the actual calculation duration, mains power supply, and photovoltaic power supply, and the optimal solution of the comprehensive power supply total cost objective function is solved. This application considers the dynamic balance between the distributed computer group and power supply costs, enabling the computer group to meet the requirements of various types of computing tasks while minimizing the comprehensive power supply total cost. While ensuring power supply stability, it maximizes the utilization of photovoltaic power and reduces mains power consumption costs.
[0019] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description
[0020] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.
[0021] Figure 1 This schematically illustrates a flowchart of the steps in an optimized scheduling method for a photovoltaic and grid-connected power supply system in an exemplary embodiment of this application. Figure 2 A schematic diagram illustrating the solar power curve of a photovoltaic system in an exemplary embodiment of this application; Figure 3 A schematic diagram illustrating the first relationship curve in an exemplary embodiment of this application; Figure 4 This is a schematic diagram illustrating the second relationship curve in an exemplary embodiment of this application; Figure 5 This diagram illustrates the cost algorithm parameters in an exemplary embodiment of this application. Figure 6 This diagram schematically illustrates the cost results of a calculation example in an exemplary embodiment of this application. Figure 7 This illustration schematically shows a comparison of the total computation time of executing the optimization strategy and not executing the method in an exemplary embodiment of this application; Figure 8 This diagram illustrates a comparison of power supply costs between implementing the optimization method and not implementing the method in an exemplary embodiment of this application. Figure 9 This diagram illustrates the power supply cost corresponding to different calculation data in an exemplary embodiment of this application. Detailed Implementation
[0022] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided to make this application more comprehensive and complete, and to fully convey the concept of the exemplary embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
[0023] Furthermore, the accompanying drawings are merely illustrative of this application and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore, repeated descriptions of them will be omitted.
[0024] This example implementation provides an optimized scheduling method for a combined photovoltaic and grid power supply system. (Reference) Figure 1 As shown, the method may include steps S101 to S108.
[0025] Step S101 involves collecting historical power generation data, historical meteorological data, computer computing power related data, and real-time meteorological data from the photovoltaic power station. The computer computing power related data includes the computer's frequency modulation range, the power consumption value corresponding to each frequency within the frequency modulation range, and the calculation speed corresponding to each power consumption value.
[0026] Step S102: Process historical meteorological data and historical power generation data to obtain a dataset.
[0027] Step S103: Divide the dataset into a training set, a validation set, and a test set. Use the training set to train the long short-term memory network model, use the validation set to validate the trained long short-term memory network model, and use the test set to evaluate the validated long short-term memory network model to obtain the photovoltaic power prediction model.
[0028] Step S104: Input real-time meteorological data into the photovoltaic power prediction model to output the photovoltaic power prediction values at each time point in the future period, and fit them into a photovoltaic daily power curve.
[0029] Step S105: Plot the relationship curve between frequency and power consumption based on the power consumption values corresponding to each frequency within the frequency modulation range, and record it as the first relationship curve. Plot the relationship curve between frequency and calculation speed based on the power consumption values corresponding to each frequency within the frequency modulation range and the calculation speed corresponding to each power consumption value, and record it as the second relationship curve. Based on the first relationship curve and the second relationship curve, obtain a set of compact power consumption values; wherein, the set of compact power consumption values can cover the power consumption value range.
[0030] Step S106: Based on the compact power consumption value and photovoltaic daily power curve of the group, determine the frequency modulation period during which the computer group needs to be frequency regulated.
[0031] Step S107: Based on the first relationship curve, the second relationship curve, and the frequency modulation time period, calculate the calculation time corresponding to each power consumption value in the group of compact power consumption values, and record the calculation time corresponding to each power consumption value in the group of compact power consumption values as the actual calculation time.
[0032] Step S108: Construct a comprehensive total power supply cost objective function based on the actual calculation time, mains power supply, and photovoltaic power supply, and solve for the optimal solution of the comprehensive total power supply cost objective function.
[0033] In one embodiment of this application, real-time meteorological data is input into a photovoltaic power prediction model using the above method to output photovoltaic power prediction values at various time points, which are then fitted into a photovoltaic daily power curve. Based on a set of compact power consumption values and the photovoltaic daily power curve, the frequency regulation time period required for the computer group can be determined. The actual calculation duration is obtained based on the first relationship curve, the second relationship curve, and the frequency regulation time period. A comprehensive power supply total cost objective function is constructed using the actual calculation duration, the mains power supply, and the photovoltaic power supply, and the optimal solution of the comprehensive power supply total cost objective function is solved. This application considers the dynamic balance between the distributed computer group and the power supply cost, enabling the computer group to meet the requirements of various types of computing tasks while minimizing the comprehensive power supply total cost. It maximizes the utilization of photovoltaic power and reduces mains power consumption costs while ensuring power supply stability.
[0034] Below, we will refer to Figures 2 to 5 The steps of the method described above in this example embodiment will be explained in more detail.
[0035] In step S101, historical power generation data, historical meteorological data, computer computing power related data, and real-time meteorological data of the photovoltaic power station are collected. The computer computing power related data includes the computer frequency modulation range, the power consumption value corresponding to each frequency within the frequency modulation range, and the calculation speed corresponding to each power consumption value.
[0036] Specifically, when collecting historical power generation data of photovoltaic power plants, the historical power generation data for historical periods is exported from the photovoltaic power plant monitoring system.
[0037] When collecting historical meteorological data for photovoltaic power plants, the specific method involves retrieving historical meteorological data for the region where the power plant is located from a public meteorological database. This historical meteorological data includes temperature, solar irradiance, air humidity, and wind speed.
[0038] When collecting computing power data for photovoltaic power plants, this data is specifically collected based on the CPU model of the computer group. Further, this computing power data includes the computer's frequency modulation range, the power consumption values corresponding to each frequency within that range, and the calculation speed corresponding to each power consumption value. Therefore, by using the CPU model, the computer's frequency modulation range, the power consumption values corresponding to each frequency within that range, and the calculation speed corresponding to each power consumption value can be obtained. The collection methods for real-time and historical meteorological data from photovoltaic power plants are the same, and the content included in real-time and historical meteorological data is identical. The difference between real-time and historical meteorological data is that real-time meteorological data is for the current time period, while historical meteorological data is for historical time periods.
[0039] In step S102, data processing includes data cleaning, data normalization and quantization, and time series construction.
[0040] Data cleaning and processing mainly includes the following: The collected historical meteorological and power generation data are cleaned to remove missing and outlier values. Linear interpolation or mean imputation methods are typically used to handle missing data.
[0041] Data normalization and quantization mainly includes the following: The input data is normalized to the [0,1] interval to avoid the impact of differences in the magnitude of feature values on model training. The normalization formula is as follows: In the formula, x' represents the normalized input data, x represents the actual measured value of the input data, min(x) is the minimum value of the input data, and max(x) is the maximum value of the input data. It should be noted that the input data includes historical meteorological data and data related to computer computing power. For example, the input data can be temperature, solar irradiance, air humidity, and wind speed.
[0042] Time series construction mainly includes the following: The input data after normalization and quantization is used to construct the input sequence using a time sliding window method: the output power of the next u+1 hours is predicted using data from the past u hours.
[0043] First, enter the sequence definition. Time series data is provided ,in This represents the feature vector at time step T. The input sequence is constructed using the sliding window method. Defined as: , Where F is the feature dimension, and R represents the set of feature parameters corresponding to time step T. The output value is then defined as follows: This represents the predicted value at time step T. The time series data D is transformed into an input-output pair using the sliding window method. The set. Specifically: The sliding window starts at time step m = T + 1 and slides sequentially until time step m = N. Input sequence and output value They satisfy a functional relationship: ,in, This represents a mapping function from the input sequence to the output value, which is fitted by a machine learning model; This is the error term.
[0044] The sliding window method described above can be used to transform time series data into input-output pairs suitable for training machine learning models, thereby constructing models for time series prediction.
[0045] Next, a model structure based on Long Short-Term Memory (LSTM) network is designed: the input shape in the input layer is (T,F); two LSTM layers are used in the LSTM layer, with 64 neurons in the first layer and 64 neurons in the second layer, to extract long-term and short-term dependency characteristics in the time series; in the fully connected layer, the output of the LSTM layer is mapped to the target value through the fully connected layer, and finally the photovoltaic power prediction value at future time points is output in the output layer.
[0046] To ensure the accuracy of this power prediction output, model training is required based on this value. A loss function based on mean squared error (MSE) is used as the calibration criterion. The expression for mean squared error is: .
[0047] in, This is the actual power value. Here, N represents the number of photovoltaic power predictions generated for the input sequence.
[0048] In steps S103 and S104, the first 70% of the dataset is divided into a training set, the middle 15% into a validation set, and the last 15% into a test set, allocated chronologically to ensure consistency between model training and validation. After the model evaluation is passed, the model receives real-time meteorological data, predicts the photovoltaic power forecast values for various time points in the future, uses the photovoltaic power forecast values as input parameters for optimized scheduling, and generates a fitted photovoltaic daily power curve, as shown in the figure. Figure 2 As shown.
[0049] It should be noted that before inputting real-time meteorological data into the photovoltaic power prediction model, data processing is required. This data processing method is the same as that used for historical meteorological data and historical power generation data. This will not be elaborated upon here.
[0050] In step S105, based on the power consumption values corresponding to each frequency within the frequency modulation range, a relationship curve between frequency and power consumption value can be plotted, i.e., the first relationship curve, as shown in Figure 105. Figure 3 As shown. Based on the power consumption values corresponding to each frequency within the frequency modulation range and the corresponding calculation speeds, a curve showing the relationship between frequency and calculation speed can be plotted, i.e., the second relationship curve. The differences in the second relationship are as follows: Figure 4 As shown. Based on the first and second relationship curves, a set of compact power consumption values can be obtained; wherein, this set of compact power consumption values can cover the power consumption value range.
[0051] It should be noted that a set of compact power consumption values includes several compact power consumption values, which are values taken from the power consumption value range according to the power consumption step size. Based on the power consumption values corresponding to each frequency within the frequency modulation range, the power consumption value range corresponding to the power consumption value can be obtained.
[0052] In step S106, based on the compact power consumption value and the photovoltaic daily power curve, the frequency modulation period for which the computer group needs to be frequency regulated can be determined.
[0053] Furthermore, step S106 includes the following: like Figure 5 As shown, by using this set of compact power consumption values and photovoltaic daily power curves, the intersection expression of the two is obtained. Solving the intersection expression yields the frequency modulation time. , The frequency modulation time period is obtained. ~ The intersection point expression is as follows: (1) In the formula, This indicates the compact power consumption value of the group. The expression representing the daily power curve of a photovoltaic system.
[0054] It is understandable that the start time of frequency modulation can be obtained by solving formula (1). and the end time of frequency modulation and obtain the frequency modulation time period ~ .
[0055] In step S107, the calculation time corresponding to each power consumption value in the group of compact power consumption values is calculated based on the first relationship curve, the second relationship curve and the frequency modulation time period, and the calculation time corresponding to each power consumption value in the group of compact power consumption values is recorded as the actual calculation time.
[0056] Furthermore, step S107 includes the following: Based on the first relationship curve, the frequency corresponding to each value in this set of compact power consumption values is obtained; Based on the second relationship curve, when the frequency is adjusted to the maximum value, the maximum computing speed corresponding to the computer group is obtained, and the shortest computing time is determined according to the maximum computing speed.
[0057] It should be noted that after determining the minimum computation time, during the calculation of the actual computation time, when the computer group is running at its maximum computation speed, the calculation can be divided into the following three cases based on whether the time of completion of the computation task falls within the frequency adjustment period: Scenario 1: When the computer group is running at maximum computing speed, the time to complete the computing task is between 0 and 12. The actual computation time is obtained based on the shortest computation time. That is, as the power consumption of the computer group gradually decreases, its corresponding computation speed also decreases, thus the actual computation time for the computer group to complete the computation task is calculated. In the shortest calculation time The calculation time is gradually increased based on the previous calculation; the expression for the actual calculation time is as follows: (2) In the formula, Indicates the actual calculation time. Indicates the shortest calculation time. This indicates that the initial power consumption of the computer group has dropped to [a certain value]. The increase in the shortest calculation time is recorded as the increase in time. The total relative workload of the computer group is obtained based on the maximum computing speed and the shortest computing time; the expression for the total relative workload of the computer group is as follows: (3) In the formula, This represents the total relative workload of the computer group. Indicates the maximum calculation speed; The actual calculation of the duration needs to be based on the increase in duration. The size is discussed in different cases, due to the increase in duration. The size is unknown beforehand, therefore the actual calculation time cannot be determined. Whether it extends to the next frequency modulation period, so the actual calculation duration The initial increase in size is based on the assumed duration. The calculation does not extend to the next frequency modulation period: when At that time, the increase in duration The frequency adjustment period does not extend to the next frequency adjustment period. At this point, based on the maximum computing speed and the frequency adjustment period, the hypothetical relative workload of the computer group within the frequency adjustment period is obtained. The expression for the hypothetical relative workload within the frequency adjustment period is as follows: (4) In the formula, This represents the hypothetical relative workload within the frequency modulation period, and n represents the number of full days that the computer group spans while performing the computing tasks. Based on the total relative workload of the computer group and the assumed relative workload during the frequency modulation period, the assumed relative workload of the computer group during the non-frequency modulation period is obtained; the expression for the assumed relative workload during the non-frequency modulation period is as follows: (5) In the formula, This represents the assumed relative workload during non-frequency modulation periods; Based on the assumed relative workload and maximum computing speed during the frequency modulation period, and the assumed relative workload and computing speed during the non-frequency modulation period, the assumed computation time is obtained; the expression for the assumed computation time is as follows: (6) In the formula, This indicates the assumed calculation time. This indicates the calculation speed during non-frequency modulation periods; After calculating the assumed calculation duration, determine whether the assumed calculation duration extends into the next frequency modulation time period; like Then the actual calculation time is equal to the assumed calculation time, that is ; like ,Right now If the calculation time has been extended into the next frequency modulation period, but the time period in which the calculation task is completed is unknown when frequency modulation is added, then the following judgment condition is added: when If the time point for completing the calculation task falls within the frequency modulation time period, then the expression for the actual calculation duration is as follows; (7) In the formula, This indicates the actual relative workload within the frequency modulation period; when If the time point for completing the calculation task falls within the period after the end of the frequency modulation period, then the expression for the actual calculation duration is as follows: (8).
[0058] Scenario 2: When the computer group operates at its maximum computing speed, the time it takes to complete the computing task falls within the frequency adjustment period. Similarly, as the power consumption of the computer group gradually decreases, its corresponding computing speed also decreases. Therefore, the actual computing time for the computer group to complete the computing task will vary. In the shortest calculation time The actual calculation time is the same as that of formula (2), and is gradually increased on the basis of the previous calculation. A time parameter is introduced and used as the final calculation time. The final calculation time is the calculation time on the last day of the shortest calculation time for the computer group to complete the calculation task at maximum computing speed. The final calculation time is obtained based on the shortest calculation time; the expression for the final calculation time is as follows: , (9) In the formula, Indicates the final calculation duration; Actual calculation time The calculations still follow the assumption principle, assuming that as the computing speed of the computer group decreases, the actual computing time for the computer group to complete the computing task, even after the increase, will still remain within the frequency modulation range. when Then, the expression for the assumed relative workload within the frequency modulation time period is as follows: (10) At this point, suppose the expression for calculating the duration is as follows: (11) like ,Right now Then the actual calculation time is equal to the assumed calculation time, that is ; like ,Right now The expression for the actual relative workload within the frequency modulation period is as follows: (12) The expression for the actual calculation time is as follows: (13).
[0059] Scenario 3: When the computer group is running at maximum computing speed, and the time point for completing the computing task is within... (During non-sunlight periods), at this time, the expression for the assumed relative workload during the frequency regulation period is the same as formula (4), the expression for the assumed relative workload during the non-frequency regulation period is the same as formula (5), and the expression for the assumed calculation duration is the same as formula (6). like ,Right now Then the actual calculation time is equal to the assumed calculation time, that is ; like ,Right now The expression for the actual relative workload within the frequency modulation time period is as follows: (14) The expression for the actual calculation time is as follows: (15) like At this time, the expression for the actual relative workload within the frequency modulation period is the same as that in formula (12), and the expression for the actual calculation duration is the same as that in formula (13).
[0060] It is understandable that when the computer group is running at its maximum computing speed, the actual computing time can be calculated in three ways depending on whether the time of completion of the computing task is within the frequency adjustment period.
[0061] In step S108, a comprehensive power supply total cost objective function is constructed based on the actual calculation time, mains power supply, and photovoltaic power supply, and the optimal solution of the comprehensive power supply total cost objective function is obtained.
[0062] It is understandable that a comprehensive power supply total cost objective function can be constructed by using the actual calculation time, mains power supply, and photovoltaic power supply. By solving this objective function, the optimal solution of the comprehensive power supply total cost objective function can be obtained. This enables the computer group to meet the requirements of various types of computing tasks while keeping the comprehensive power supply total cost to a minimum. This allows the present application to maximize the use of photovoltaic power and reduce the cost of mains power consumption while ensuring power supply stability.
[0063] In one embodiment, the daily mains power purchase amount of the integrated power supply system is obtained based on the daytime and nighttime mains power purchase amounts; wherein, the expression for the daily mains power purchase amount of the integrated power supply system is as follows: (16) In the formula, This represents the daily amount of mains electricity purchased by the integrated power supply system. This indicates the amount of electricity purchased during the day. This indicates the amount of electricity purchased at night. When the maximum power consumption value in this set of compact power consumption values is greater than or equal to the maximum daily photovoltaic power, the expression for the daytime grid electricity purchase quantity is as follows: (17) In the formula, This indicates the maximum power consumption value within this group of compact power consumption values. This indicates the maximum daily power output of photovoltaic power. The value is 6:00, which is recorded as 0 in integral calculations. It is 18:00, which is recorded as 13 in integral calculation; When the maximum power consumption value in this set of compact power consumption values is less than the maximum daily photovoltaic power, by Solve for the starting time point when the computer group is powered solely by photovoltaics. and the end time of photovoltaic power supply The period when only photovoltaic power is supplied is ~ ,exist ~ The expression for daytime electricity purchases is as follows: (18) The expression for nighttime electricity purchases is as follows: (19).
[0064] It is understandable that the daily mains power purchase amount of the integrated power supply system can be calculated using formula (16). Specifically, when the maximum power consumption value in this set of compact power consumption values is greater than or equal to the maximum daily photovoltaic power, the daytime mains power purchase amount can be calculated using formula (17). When the maximum power consumption value in this set of compact power consumption values is less than the maximum daily photovoltaic power, the daytime mains power purchase amount can be calculated using formula (18). The nighttime mains power purchase amount can be calculated using formula (19).
[0065] In one embodiment, the photovoltaic power supply cost is calculated based on the photovoltaic power generation and the levelized cost of electricity (LCOE) of the photovoltaic power plant; wherein, the expression for the daily photovoltaic power supply cost of the integrated power supply system is as follows: (20) In the formula, This indicates the cost per kilowatt-hour of a photovoltaic power plant. This represents the daily photovoltaic power supply of the integrated power supply system. This indicates the daily photovoltaic power supply cost of the integrated power supply system.
[0066] It is understandable that the daily photovoltaic power supply cost of the integrated power supply system can be calculated using formula (20).
[0067] In one embodiment, the expression for the total cost of electricity supply over n full days is as follows: (twenty one) In the formula, This represents the total cost of electricity supply over n complete days. This represents the daily cost of mains electricity purchased by the integrated power supply system, where 's' is the unit price of electricity purchased from the main grid. This indicates the levelized cost of electricity (LCOE) of a photovoltaic system. The expression for the actual calculation time is as follows: (twenty two) When the maximum power consumption value in this set of compact power consumption values Greater than or equal to the maximum daily photovoltaic power The expression for the cost of power supply during the remaining time is as follows: (twenty three) When the maximum power consumption value in this set of compact power consumption values Less than the maximum daily power of photovoltaic power The expression for the cost of power supply during the remaining time is as follows: (twenty four) In the formula, This indicates the cost of supplying electricity for the remaining duration.
[0068] Understandably, the comprehensive power supply cost over n full days can be calculated using formula (21). The final calculation time can be calculated using formula (22) and the actual calculation time obtained in the previous three scenarios. .
[0069] When the maximum power consumption value in this set of compact power consumption values Greater than or equal to the maximum daily photovoltaic power At that time, based on the final calculated duration Is it during the FM tuning period? Within this timeframe, calculate the cost of electricity supply for the remaining duration. Specifically, the power supply cost for the remaining duration can be calculated using formula (23). .
[0070] When the maximum power consumption value in this set of compact power consumption values Less than the maximum daily power of photovoltaic power At that time, based on the final calculated duration Is it during the FM tuning period? Inside, and the final calculation duration The starting time of photovoltaic power supply The end time of photovoltaic power supply The relationship between the magnitudes is discussed in different cases, that is, determining the specific time period in which the calculation task ends, as shown in formula (24), to calculate the power supply cost within the remaining time. Specifically, the power supply cost for the remaining duration can be calculated using formula (24). .
[0071] In one embodiment, the objective function for the total cost of power supply is expressed as follows: (25) In the formula, This represents the objective function for the total cost of integrated power supply.
[0072] It is understandable that the objective function for the total cost of integrated power supply can be obtained through formula (25). The integrated power supply cost over n complete days calculated using formula (21) is then used. And the cost of supplying electricity during the remaining duration calculated by formula (23) or formula (24). Substituting this into formula (3), the corresponding overall total cost objective function can be obtained, and the optimal solution can be obtained. Thus, the computer group can meet the requirements of various types of computing tasks while keeping the overall power supply cost to a minimum, so that this application can maximize the use of photovoltaic power and reduce the cost of mains power consumption while ensuring power supply stability.
[0073] The following numerical examples further illustrate this application.
[0074] Calculation example 1: For example, the installed capacity of a photovoltaic power station is 10kW, and the CPU model of the computer group is AMD Ryzen 9 3950X, with a frequency modulation range of 2.5~3.9GHz. The corresponding power consumption values within this range are shown in Table 1.
[0075] Table 1. Frequency and Corresponding Power Consumption Values Within the frequency modulation range, the changes in frequency and power consumption are observed by adjusting the load, and the frequency and power consumption under steady-state conditions are recorded to plot the first relationship curve.
[0076] Adjust the frequency of the computer group and record the power consumption value corresponding to each frequency. Record the computation time corresponding to each power consumption value under identical load conditions. Determine the shortest computation time when the computer group frequency is adjusted to its maximum, and obtain the computation speed corresponding to each power consumption value. Then, plot the second relationship curve based on frequency and computation speed. The first relationship curve is shown below. Figure 3As shown, the second relationship curve is as follows Figure 4 As shown.
[0077] Based on the first and second relationship curves, a set of compact power consumption values is obtained. According to the power consumption range determined in Table 1, a set of compact power consumption values is selected with a power consumption step size of 0.2kW: = (6, 5.8, 5.6, 5.4, 5.2, 5.0, 4.8, 4.6, 4.4, 4.2, 4.0, 3.8, 3.6, 3.4, 3.2, 3.0).
[0078] Retrieve historical power generation data and historical meteorological data from a photovoltaic power station over the past summer. Construct a time series data D from the acquired historical power generation data and historical meteorological data using a time sliding window method, and then convert the time series data D into an input-output pair. The dataset is divided into training, validation, and test sets. The training set is used to train the Long Short-Term Memory (LSTM) network model, the validation set is used to validate the trained LSM network model, and the test set is used to evaluate the validated LSM network model. Once the model passes the evaluation, a photovoltaic (PV) power prediction model is obtained. The PV power prediction model is input with processed real-time meteorological data, and finally, the output layer outputs the predicted PV power values for each time point in the future period, such as... Figure 2 As shown, the expression for the fitted photovoltaic daily power curve is: t represents a point in time. First, determine the frequency modulation time period according to formula (1). ~ .
[0079] The calculation results are shown in Table 2.
[0080] Table 2 shows the frequency modulation periods corresponding to each of the compact power consumption values. The shortest computation time to complete a computing task when the computer group reaches its maximum frequency. From this, we can know the total relative workload of the computer group. Furthermore, when the computer is running at its maximum computing speed, the time to complete the computing task is between 0 and... .
[0081] Assumed relative workload within the frequency modulation period: ; Hypothetical relative workload during non-frequency modulation periods: ; Assuming the calculation time is calculated according to formula (6), the calculation results of the calculation time are shown in Table 3.
[0082] The above shows the assumed calculation time corresponding to power consumption values of 3.2kW and 3kW, and the corresponding increase in calculation time. As the frequency regulation cycle increases, the actual calculation time corresponding to the power consumption values of 3.2kW and 3kW calculated according to formulas (7) and (8) is 27.896h and 30.1445h, respectively.
[0083] Once the actual calculation time is determined, the total cost of combined power supply is calculated.
[0084] First, calculate the daily photovoltaic power supply of the integrated power supply system. The daily photovoltaic power supply of the integrated power supply system The calculation is as follows: Yuan.
[0085] Then calculate the daily mains power purchase cost of the integrated power supply system. Calculate the daily mains power purchase cost of the integrated power supply system. First, calculate the daily electricity purchase volume of the integrated power supply system. Daily electricity purchases by the integrated power supply system The calculation is as follows: Daily mains power purchase cost of integrated power supply system The calculation is as follows: .
[0086] Next, the final calculation time corresponding to each power consumption value. They are not the same. The power supply cost for the computer group on the last day (i.e., the power supply cost for the remaining time) is calculated according to formula (19). For example... Figure 5 As shown, the final power supply cost results for all time periods are summed to obtain the power supply cost result for the computer group on the last day.
[0087] This application compares the calculation results of the optimized scheduling in the instance with the calculation results of the same instance but without implementing the optimized scheduling method. Clearly, in the instance without implementing the optimized scheduling method, no frequency regulation operation was applied, resulting in a significant waste of electrical energy by the photovoltaic power station during sunshine hours. This energy was not consumed in a timely manner, greatly weakening the economic efficiency of the power supply scheme. Figure 8 As shown in the figure, this only considers that the cost of photovoltaic electricity remains constant, without taking into account a series of subsequent costs such as equipment depreciation, maintenance costs, and even the cost of energy storage equipment caused by photovoltaic curtailment. Meanwhile, the computation time of computer groups after optimized scheduling is generally shorter, and this advantage becomes increasingly apparent as the computational workload increases. Figure 7 .
[0088] Although the above Figure 7 The results in the original calculation were obtained using data from a specific example and the algorithm of this optimization method. However, to demonstrate that the method is also applicable to other example data, the installed capacity of the photovoltaic power plant was expanded from the original fixed value of 10kW to... This variable range, according to the optimized scheduling method, calculates the power supply cost data set corresponding to each value within this range, and generates data such as... Figure 9 The three-dimensional data diagram shows that, although the computing power data of the computer group did not change during the process, the general relationship between the output power of the photovoltaic power station and the power consumption of the unit was included, which shows the universality of this method in the field of comprehensive energy system optimization and scheduling for computer group power consumption.
[0089] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.
Claims
1. An optimized scheduling method for a photovoltaic and grid-connected power supply system, characterized in that, include: Collect historical power generation data, historical meteorological data, computer computing power related data, and real-time meteorological data of photovoltaic power plants; the computer computing power related data includes the computer frequency modulation range, the power consumption value corresponding to each frequency within the frequency modulation range, and the calculation speed corresponding to each power consumption value; Data processing is performed on the historical meteorological data and the historical power generation data to obtain a dataset; The dataset is divided into a training set, a validation set, and a test set. The long short-term memory network model is trained using the training set, the trained long short-term memory network model is validated using the validation set, and the validated long short-term memory network model is evaluated using the test set to obtain the photovoltaic power prediction model. The real-time meteorological data is input into the photovoltaic power prediction model to output the photovoltaic power prediction values at each time point in the future period, and then fitted into a photovoltaic daily power curve. A relationship curve between frequency and power consumption value is plotted based on the power consumption value corresponding to each frequency within the frequency modulation range, and is denoted as the first relationship curve. A relationship curve between frequency and calculation speed is plotted based on the power consumption value corresponding to each frequency within the frequency modulation range and the calculation speed corresponding to each power consumption value, and is denoted as the second relationship curve. A set of compact power consumption values is obtained based on the first relationship curve and the second relationship curve. The set of compact power consumption values can cover the power consumption value range. Based on the compact power consumption value and the photovoltaic daily power curve, the frequency modulation period for which the computer group needs to be frequency-modulated is determined; Based on the first relationship curve, the second relationship curve, and the frequency modulation time period, calculate the calculation time corresponding to each power value in the group of compact power consumption values, and record the calculation time corresponding to each power value in the group of compact power consumption values as the actual calculation time; Based on the actual calculation time, mains power supply, and photovoltaic power supply, a comprehensive total power supply cost objective function is constructed, and the optimal solution of the comprehensive total power supply cost objective function is obtained.
2. The optimized scheduling method for a photovoltaic and grid-connected power supply system according to claim 1, characterized in that, The step of determining the frequency modulation period for which the computer group needs frequency modulation based on the compact power consumption value of the group and the photovoltaic daily power curve includes: By combining the compact power consumption values and the photovoltaic daily power curve, the intersection expression of the two is obtained. Solving the intersection expression yields the frequency modulation start time. and the end time of frequency modulation The frequency modulation time period is ~ The intersection point expression is as follows: （1） In the formula, This indicates the compact power consumption value of the group. The expression representing the daily power curve of a photovoltaic system.
3. The optimized scheduling method for a photovoltaic and grid-connected power supply system according to claim 2, characterized in that, The step of calculating the calculation time corresponding to each power consumption value in the set of compact power consumption values based on the first relationship curve, the second relationship curve, and the frequency modulation time period, and recording the calculation time corresponding to each power consumption value in the set of compact power consumption values as the actual calculation time, includes: Based on the first relationship curve, the frequency corresponding to each value in this set of compact power consumption values is obtained; Based on the second relationship curve, when the frequency is adjusted to the maximum value, the maximum computing speed corresponding to the computer group is obtained, and the shortest computing time is determined according to the maximum computing speed.
4. The optimized scheduling method for a photovoltaic and grid-connected power supply system according to claim 3, characterized in that, The step of calculating the calculation time corresponding to each power consumption value in the set of compact power consumption values based on the first relationship curve, the second relationship curve, and the frequency modulation time period, and recording the calculation time corresponding to each power consumption value in the set of compact power consumption values as the actual calculation time, includes: When the computer group is running at maximum computing speed, the time point for completing the computing task is between 0 and 10. The actual calculation time is obtained based on the shortest calculation time; wherein, the expression for the actual calculation time is as follows: （2） In the formula, Indicates the actual calculation time. Indicates the shortest calculation time. This indicates that the initial power consumption of the computer group has dropped to [a certain value]. The increase in the shortest calculation time is recorded as the increase in time. Based on the maximum computing speed and the shortest computing time, the total relative workload of the computer group is obtained; wherein, the expression for the total relative workload of the computer group is as follows: （3） In the formula, This represents the total relative workload of the computer group. Indicates the maximum calculation speed; when At that time, based on the maximum computing speed and the frequency modulation time period, the assumed relative workload of the computer group within the frequency modulation time period is obtained; wherein, the expression for the assumed relative workload within the frequency modulation time period is as follows: （4） In the formula, This represents the hypothetical relative workload within the frequency modulation period, and n represents the number of full days that the computer group spans while performing the computing tasks. Based on the total relative workload of the computer group and the assumed relative workload during the frequency modulation period, the assumed relative workload of the computer group during the non-frequency modulation period is obtained; wherein, the expression for the assumed relative workload during the non-frequency modulation period is as follows: （5） In the formula, This represents the assumed relative workload during non-frequency modulation periods; Based on the assumed relative workload during the frequency modulation period, the maximum computing speed, the assumed relative workload during the non-frequency modulation period, and the computing speed during the non-frequency modulation period, the assumed computing time is obtained; wherein, the expression for the assumed computing time is as follows: （6） In the formula, This indicates the assumed calculation time. This indicates the calculation speed during non-frequency modulation periods; Determine whether the assumed calculation duration extends into the next frequency modulation time period; like Then the actual calculation time is equal to the assumed calculation time, that is ; like ,Right now Then it is assumed that the calculation time has been extended into the next frequency modulation time period; when If the time point for completing the calculation task falls within the frequency modulation time period, then the expression for the actual calculation duration is as follows; （7） In the formula, This indicates the actual relative workload within the frequency modulation period; when If the time point for completing the calculation task falls within the time period after the end of the frequency modulation period, then the expression for the actual calculation duration is as follows: （8）。 5. The optimized scheduling method for a photovoltaic and grid-connected power supply system according to claim 4, characterized in that, The step of calculating the calculation time corresponding to each power consumption value in the set of compact power consumption values based on the first relationship curve, the second relationship curve, and the frequency modulation time period, and recording the calculation time corresponding to each power consumption value in the set of compact power consumption values as the actual calculation time, includes: When the computer group is running at maximum computing speed, the time point for completing the computing task is within the frequency modulation time period, and the expression for the actual computing time is the same as formula (2); The final computation time is obtained based on the shortest computation time; wherein, the final computation time is the computation time on the last day of the shortest computation time for the computer group to complete the computation task at the maximum computation speed, and the expression for the final computation time is as follows: ， （9） In the formula, Indicates the final calculation duration; when Then, the expression for the assumed relative workload within the frequency modulation time period is as follows: （10） At this point, the expression for the assumed calculation time is as follows: （11） like ,Right now Then the actual calculation time is equal to the assumed calculation time, that is ; like ,Right now The expression for the actual relative workload within the frequency modulation time period is as follows: （12） The expression for the actual calculation time is as follows: （13）。 6. The optimized scheduling method for a photovoltaic and grid-connected power supply system according to claim 5, characterized in that, The step of calculating the calculation time corresponding to each power consumption value in the set of compact power consumption values based on the first relationship curve, the second relationship curve, and the frequency modulation time period, and recording the calculation time corresponding to each power consumption value in the set of compact power consumption values as the actual calculation time, includes: When the computer group is running at maximum computing speed, and the time point for completing the computing task is located at At this time, the expression for the assumed relative workload during the frequency modulation period is the same as formula (4), the expression for the assumed relative workload during the non-frequency modulation period is the same as formula (5), and the expression for the assumed calculation duration is the same as formula (6). like ,Right now Then the actual calculation time is equal to the assumed calculation time, that is ; like ,Right now The expression for the actual relative workload within the frequency modulation time period is as follows: （14） The expression for the actual calculation time is as follows: （15） like At this time, the expression for the actual relative workload within the frequency modulation time period is the same as that of formula (12), and the expression for the actual calculation duration is the same as that of formula (13).
7. The optimized scheduling method for a photovoltaic and grid-connected power supply system according to claim 6, characterized in that, The daily mains power purchase volume of the integrated power supply system is obtained based on the daytime and nighttime mains power purchase volumes; wherein, the expression for the daily mains power purchase volume of the integrated power supply system is as follows: （16） In the formula, This represents the daily amount of mains electricity purchased by the integrated power supply system. This indicates the amount of electricity purchased during the day. This indicates the amount of electricity purchased at night. When the maximum power consumption value in this set of compact power consumption values is greater than or equal to the maximum daily photovoltaic power, the expression for the daytime grid electricity purchase quantity is as follows: （17） In the formula, This indicates the maximum power consumption value within this group of compact power consumption values. This indicates the maximum daily power output of photovoltaic power. The value is 6:00, which is recorded as 0 in integral calculations. It is 18:00, which is recorded as 13 in integral calculation; When the maximum power consumption value in this set of compact power consumption values is less than the maximum daily photovoltaic power, by Solve for the start time of photovoltaic power supply and the end time of photovoltaic power supply The period during which only the photovoltaic power supply is ~ ,exist ~ The expression for the daytime electricity purchase volume is as follows: （18） The expression for the nighttime electricity purchase volume is as follows: （19）。 8. The optimized scheduling method for a photovoltaic and grid-connected power supply system according to claim 7, characterized in that, The photovoltaic power supply cost is calculated based on the photovoltaic power generation and the levelized cost of electricity (LCOE) of the photovoltaic power station; the expression for the daily photovoltaic power supply cost of the integrated power supply system is as follows: （20） In the formula, This indicates the cost per kilowatt-hour of a photovoltaic power plant. This represents the daily photovoltaic power supply of the integrated power supply system. This indicates the daily photovoltaic power supply cost of the integrated power supply system.
9. The optimized scheduling method for a photovoltaic and grid-connected power supply system according to claim 8, characterized in that, The expression for the comprehensive cost of electricity supply over n full days is as follows: （21） In the formula, This represents the total cost of electricity supply over n complete days. This represents the daily cost of mains electricity purchased by the integrated power supply system, where 's' is the unit price of electricity purchased from the main grid. This indicates the cost per kilowatt-hour of a photovoltaic power plant; The expression for the actual calculation time is as follows: （22） When the maximum power consumption value in this set of compact power consumption values is greater than or equal to the maximum daily photovoltaic power, the expression for the power supply cost over the remaining time is as follows: （23） When the maximum power consumption value in this set of compact power consumption values is less than the maximum daily photovoltaic power, the expression for the power supply cost during the remaining time is as follows: （24） In the formula, This indicates the cost of supplying electricity for the remaining duration.
10. The optimized scheduling method for a photovoltaic and grid-connected power supply system according to claim 9, characterized in that, The expression for the overall power supply cost objective function is as follows: （25） In the formula, This represents the objective function for the total cost of integrated power supply.