Source network load storage integrated collaborative control method and device, electronic equipment and medium

Abstract

The present application relates to source network load storage collaborative control technical field, especially to a kind of source network load storage integrated collaborative control method, device, electronic equipment and medium, the present application method first obtains first historical load data queue;Then the periodicity inspection is carried out to first historical load data queue, and first smoothing queue is obtained;Then data is extracted from first smoothing queue and input to load prediction equation, and first load prediction data queue is obtained;Finally, according to photovoltaic expected output data queue and first load prediction data queue, according to the first target equation optimization energy storage power, obtains first energy storage power data queue.The present application improves energy utilization efficiency and economy through multi-objective collaborative optimization.Taking the minimum of abandoned energy, photovoltaic output fluctuation, energy storage device charge-discharge amount as the core target, combined with energy storage operation constraint to build the first target equation and optimize energy storage power, the balanced promotion of multi-objective is realized.

Description

Technical Field

[0001] This invention relates to the field of source-grid-load-storage coordinated control technology, and in particular to an integrated source-grid-load-storage coordinated control method, device, electronic equipment and medium. Background Technology

[0002] The energy system is accelerating its transition to low-carbon and zero-carbon, and the integrated generation, grid, load, and energy storage system has become a core component of the new power system. By integrating resources from the power source (renewable energy sources such as photovoltaics), the grid, the load, and energy storage, it achieves multi-source coordinated dispatch, which is a key path to improve energy efficiency and ensure the safe and stable operation of the power system. Among these, the accuracy of load forecasting and the effectiveness of energy storage power optimization directly determine the operational efficiency of the generation-grid-load-storage grid structure. Especially after the large-scale integration of renewable energy sources such as photovoltaics, their output is significantly affected by natural factors such as sunlight and weather, further exacerbating the difficulty of load forecasting and the complexity of energy storage dispatch, thus placing higher demands on load forecasting and energy storage optimization technologies.

[0003] Currently, in the actual operation of the power grid structure of source-grid-load-storage, there are still many technical deficiencies in load forecasting and energy storage power optimization, which make it difficult to meet the requirements of efficient, economical and safe operation of the system. The specific deficiencies are as follows: First, the basic data processing is not standardized, resulting in insufficient reliability of subsequent technical steps. In existing technologies, there is a lack of unified standards for collecting historical load data before load forecasting, which often fails to fully guarantee the continuity and integrity of the data. This can easily lead to problems such as missing data and outlier interference, causing deviations in the input data for subsequent periodic checks, load forecasting, and energy storage optimization. This affects the reliability of the entire process from the source and fails to provide high-quality data support for subsequent technical steps.

[0004] Secondly, the volatility of load data is not effectively addressed, making it difficult to improve forecast accuracy. Load data in the power grid structure of source-grid-load-storage is affected by periodic factors such as daily and weekly cycles, exhibiting obvious non-stationary characteristics. However, existing load forecasting methods mostly use raw non-stationary data directly for model construction without addressing the periodic fluctuations of the data. This results in forecasting models failing to accurately capture load change patterns, leading to significant forecast biases and an inability to truly reflect future load changes. Consequently, this affects the rationality of subsequent energy storage optimization and system scheduling.

[0005] Third, load forecasting models have poor adaptability and lack adaptive capabilities. Existing load forecasting models mostly use fixed parameters and a single structure, without adaptively adjusting to the dynamic characteristics of load data. They cannot adapt to different time scales and load fluctuation characteristics in the power grid structure of source-grid-load-storage, which easily leads to insufficient model adaptability, further exacerbating forecasting errors, causing scheduling decision errors, and affecting system operating efficiency.

[0006] Fourth, the optimization objectives of energy storage power are singular, failing to achieve a coordinated balance among multiple objectives. Existing energy storage power optimization schemes mostly focus on a single objective (such as minimizing energy curtailment or minimizing energy storage charging and discharging losses), without comprehensively considering the needs of multiple aspects such as the volatility of photovoltaic output and the charging and discharging capacity of energy storage devices. This results in limitations in optimization effectiveness—either excessive photovoltaic energy curtailment leading to low energy utilization efficiency; or excessive fluctuations in photovoltaic output impacting grid stability; or excessively frequent charging and discharging of energy storage devices, exacerbating equipment wear and shortening service life, failing to achieve coordinated optimization of all aspects of the power generation, grid, load, and storage system.

[0007] In summary, current load forecasting and energy storage power optimization technologies in the power grid structure of source-grid-load-storage still suffer from technical defects such as non-standard data processing, low forecasting accuracy, poor model adaptability, insufficient multi-objective coordination, imperfect energy storage constraints, and fragmented processes. These defects fail to meet the actual needs of efficient, economical, and safe operation of integrated source-grid-load-storage systems. Therefore, there is an urgent need for a load forecasting and energy storage optimization solution that can solve the above-mentioned technical problems. Summary of the Invention

[0008] The present invention provides a method, device, electronic device and medium for integrated source-grid-load-storage control, which is used to solve the problem that it is difficult to achieve multi-objective coordination in the existing integrated source-grid-load-storage system.

[0009] In a first aspect, embodiments of the present invention provide an integrated source-grid-load-storage coordinated control method, comprising: Obtain the first historical load data queue; The first historical load data queue is periodically tested, and the first historical load data queue is differentiated based on the first period length obtained from the test to obtain the first stable queue. A load forecasting equation is constructed based on the first stabilization queue, and data is extracted from the first stabilization queue and input into the load forecasting equation to obtain the first load forecasting data queue. Based on the photovoltaic expected output data queue and the first load prediction data queue, a first objective equation is constructed with the goal of minimizing abandoned energy, the volatility of photovoltaic output, and the charging and discharging capacity of energy storage devices, and constrained by the operating status of energy storage devices. The energy storage power is optimized based on the first objective equation to obtain the first energy storage power data queue.

[0010] In one possible implementation, the step of periodically inspecting the first historical load data queue and differentially processing the first historical load data queue based on the first period length obtained from the inspection to obtain a first stable queue includes: Obtain the length of the second cycle; Based on the second cycle length, data pairs are extracted from the first historical load data queue in a sliding manner, and the extracted data pairs are added to the data pair queue. The periodic index is calculated for the queue based on the data, and the calculation result is added to the periodic index queue. If the second cycle length does not reach the threshold, the second cycle length is incremented or decremented sequentially, and the process jumps to the step of extracting data pairs from the first historical load data queue in a sliding manner according to the second cycle length, and adding the extracted data pairs to the data pair queue. Otherwise, the second period length corresponding to the minimum value in the periodic index queue is taken as the first period length, and the data pair queue corresponding to the first period length is taken as the target data pair queue. Extract data pairs sequentially from the target data pair queue; For each extracted data pair, perform a difference operation in a predetermined order to obtain the difference value; Based on the position of the data pair in the target data queue, multiple difference values ​​are constructed into the first stabilization queue.

[0011] In one possible implementation, the step of calculating a periodic index on the queue based on the data and adding the calculated result to the periodic index queue includes: The periodic index is calculated for the queue based on the first formula and the data, and the calculated result is added to the periodic index queue. The first formula is:

[0012] In the formula, It is a cyclical index. This represents the total number of data pairs in the data pair queue. For the data pair queue, the first The second value in a pair of data. For the data pair queue, the first The first value of a pair of data.

[0013] In one possible implementation, the step of constructing a load forecasting equation based on the first smoothing queue and extracting data from the first smoothing queue and inputting it into the load forecasting equation to obtain a first load forecasting data queue includes: Obtain the first fundamental time series equation and multiple first lengths; For each first length, multiple first data segments and multiple reference data are extracted from the first stabilization queue in a sliding manner according to the first length, and the extracted multiple first data segments and multiple reference data are constructed into a first dataset, wherein each first data segment corresponds to a reference data, and the reference data corresponding to the first data segment is connected to the first data segment in the first stabilization queue; For each first dataset, the number of parameters of the basic time series equation is adjusted according to the first length corresponding to the first dataset to obtain the second basic time series equation; Prediction bias acquisition steps: For each second basic time series equation, substitute multiple first data segments from the first dataset corresponding to the second basic time series equation into the second basic time series equation, and determine the prediction bias of the second basic time series equation based on the multiple outputs obtained and multiple reference data from the first dataset corresponding to the second basic time series equation. If there is a prediction deviation less than the deviation threshold among multiple prediction deviations, then the second basic time series equation corresponding to the smallest prediction deviation is taken as the load prediction equation, and the first length corresponding to the load prediction equation is taken as the equation data length. Otherwise, based on multiple prediction biases, the parameters of multiple second basic time series equations are adjusted, and the process jumps to the prediction bias acquisition step. Based on the length of the equation data, multiple second data segments are extracted from the first stabilization queue in a sliding manner; The multiple second data segments are input into the load prediction equation to obtain multiple differential prediction values; For each differential prediction value, the target historical load is extracted from the first historical load data queue based on the first cycle length and the differential prediction value; For each target historical load, the sum of the target historical load and the differential prediction value is added to the first load prediction data queue as load prediction data.

[0014] In one possible implementation, the first basic timing equation is:

[0015] In the formula, These are the difference prediction values. For the first The weight parameters for each input difference value. For the first Each input difference value For the intercept parameter, This is the total number of input difference values.

[0016] In one possible implementation, the construction of a first objective equation based on the photovoltaic expected output data queue and the first load forecast data queue, with the objectives of minimizing energy curtailment, photovoltaic output volatility, and the charging and discharging capacity of the energy storage device, and constrained by the operating state of the energy storage device, includes: Based on the first load forecast data queue, a power balance equation is constructed to characterize the relationship between photovoltaic power output, energy storage device output, main power output, and load power. Based on the photovoltaic expected output data queue, construct the curtailment equation characterizing the relationship between photovoltaic output and curtailed energy, and the fluctuation equation characterizing the relationship between photovoltaic output and photovoltaic output volatility. Construct a charge / discharge equation to characterize the charge / discharge rate of an energy storage device; Different weighting coefficients are assigned to the curtailment of energy, the volatility of photovoltaic power output, and the charging and discharging capacity of energy storage devices, and a second objective equation is constructed based on the weighting coefficients; Construct constraint equations relating the state of charge (SBC) of an energy storage device to its charge and discharge limits; The first objective equation is obtained by combining the constraint equation with the second objective equation.

[0017] In one possible implementation, the first objective equation is:

[0018] In the formula, for Load power at time points, for Photovoltaic output at specific time points for Main power output at the time point The charging and discharging power of the energy storage device, To discard energy, This represents the total number of time points predicted for the first load forecast data queue. for Expected photovoltaic output at the specified time point The charge and discharge capacity of the energy storage device. This is the second objective equation. , as well as These are the first weighting coefficient, the second weighting coefficient, and the third weighting coefficient, respectively. for The state of charge of the energy storage device at a given time point. for The state of charge of the energy storage device at a given time point. The charging power conversion factor is... Rated reference power, This is the discharge smoothness coefficient. This is the charging smoothness coefficient. Function to set negative values ​​to zero.

[0019] Secondly, embodiments of the present invention provide an integrated source-grid-load-storage coordinated control device for implementing the integrated source-grid-load-storage coordinated control method as described in the first aspect or any possible implementation thereof, the integrated source-grid-load-storage coordinated control device comprising: The historical load acquisition module is used to acquire the first historical load data queue; The load data processing module is used to periodically check the first historical load data queue and perform differential processing on the first historical load data queue according to the first period length obtained by the check to obtain a first stable queue. The load forecasting module is used to construct a load forecasting equation based on the first stabilization queue, and extract data from the first stabilization queue and input it into the load forecasting equation to obtain a first load forecasting data queue. as well as, The integrated control module is used to construct a first objective equation based on the photovoltaic expected output data queue and the first load prediction data queue, with the goal of minimizing the curtailment of energy, the volatility of photovoltaic output and the charging and discharging of energy storage devices, and constrained by the operating status of energy storage devices. The module optimizes the energy storage power based on the first objective equation and obtains the first energy storage power data queue.

[0020] Thirdly, embodiments of the present invention provide an electronic device, including a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor executes the computer program to implement the steps of the method as described in the first aspect or any possible implementation of the first aspect.

[0021] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in the first aspect or any possible implementation thereof.

[0022] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows: This invention discloses an integrated source-grid-load-storage coordinated control method. First, a first historical load data queue is acquired. Then, the first historical load data queue is periodically checked, and differential processing is performed on the first historical load data queue based on the first period length obtained from the checks, resulting in a first stable queue. Next, a load prediction equation is constructed based on the first stable queue, and data extracted from the first stable queue is input into the load prediction equation to obtain a first load prediction data queue. Finally, based on the photovoltaic expected output data queue and the first load prediction data queue, a first objective equation is constructed with the goal of minimizing energy curtailment, photovoltaic output volatility, and the charging and discharging capacity of the energy storage device, constrained by the operating state of the energy storage device. The energy storage power is optimized based on the first objective equation to obtain a first energy storage power data queue.

[0023] This invention improves the accuracy and applicability of load forecasting by optimizing the prediction model. The true period of the first historical load data queue is determined through periodic verification, and then a first stable queue is constructed through differential processing. This effectively eliminates periodic fluctuations (such as daily and weekly load fluctuations) in the original load data, bringing the data to a stable state. This processing method solves the technical pain point that non-stationary data is difficult to use directly for forecasting. Compared with directly using raw non-stationary data for forecasting, it significantly improves the accuracy of load forecasting, ensuring that the first load forecast data queue can truly reflect future load changes, providing accurate load reference for subsequent energy storage optimization and system scheduling. By constructing datasets with multiple sets of first-length data, adjusting the time-series equation parameters, and selecting the optimal load forecasting equation, precise adaptation between the prediction model and data characteristics is achieved.

[0024] This invention improves energy utilization efficiency and economy through multi-objective collaborative optimization. With the minimization of energy curtailment, photovoltaic power output fluctuations, and the charging and discharging capacity of energy storage devices as core objectives, a first objective equation is constructed in conjunction with energy storage operation constraints, and the energy storage power is optimized, achieving a balanced improvement across multiple objectives. On the one hand, by minimizing energy curtailment, the utilization efficiency of photovoltaic energy is effectively improved, photovoltaic energy waste is reduced, and energy losses in the power system are lowered, meeting the development needs of energy conservation and emission reduction. On the other hand, by minimizing photovoltaic power output fluctuations, the impact of photovoltaic power output fluctuations on the power system is mitigated, the regulation pressure on the main power source is reduced, the operating cost of the main power source is lowered, and the economic efficiency of system operation is improved. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a flowchart of the integrated source-grid-load-storage coordinated control method provided in the embodiments of the present invention; Figure 2 This is a schematic diagram of the first smoothing queue acquisition process provided by an embodiment of the present invention; Figure 3 This is a functional block diagram of the integrated source-grid-load-storage coordinated control device provided in the embodiments of the present invention; Figure 4 This is a functional block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0027] In the following description, specific details such as particular system structures and techniques are set forth for illustrative purposes and not for limitation, so as to provide a thorough understanding of embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0028] To make the objectives, technical solutions, and advantages of the present invention clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0029] The embodiments of the present invention will be described in detail below. This example is implemented based on the technical solution of the present invention, and provides detailed implementation methods and specific operation processes. However, the protection scope of the present invention is not limited to the following embodiments.

[0030] Figure 1 A flowchart of the integrated source-grid-load-storage coordinated control method provided for embodiments of the present invention.

[0031] like Figure 1 As shown, it illustrates the implementation flowchart of the integrated source-grid-load-storage coordinated control method provided by the embodiments of the present invention, which is described in detail below: In step 101, the first historical load data queue is obtained.

[0032] In step 102, the first historical load data queue is periodically inspected, and the first historical load data queue is differentiated according to the first period length obtained from the inspection to obtain the first stable queue.

[0033] In some implementations, the step of periodically inspecting the first historical load data queue and differentially processing the first historical load data queue based on the first period length obtained from the inspection to obtain a first stable queue includes: Obtain the length of the second cycle; Based on the second cycle length, data pairs are extracted from the first historical load data queue in a sliding manner, and the extracted data pairs are added to the data pair queue. The periodic index is calculated for the queue based on the data, and the calculation result is added to the periodic index queue. If the second cycle length does not reach the threshold, the second cycle length is incremented or decremented sequentially, and the process jumps to the step of extracting data pairs from the first historical load data queue in a sliding manner according to the second cycle length, and adding the extracted data pairs to the data pair queue. Otherwise, the second period length corresponding to the minimum value in the periodic index queue is taken as the first period length, and the data pair queue corresponding to the first period length is taken as the target data pair queue. Extract data pairs sequentially from the target data pair queue; For each extracted data pair, perform a difference operation in a predetermined order to obtain the difference value; Based on the position of the data pair in the target data queue, multiple difference values ​​are constructed into the first stabilization queue.

[0034] In some implementations, the step of calculating a periodic index on the queue based on the data and adding the calculated result to the periodic index queue includes: The periodic index is calculated for the queue based on the first formula and the data, and the calculated result is added to the periodic index queue. The first formula is:

[0035] In the formula, It is a cyclical index. This represents the total number of data pairs in the data pair queue. For the data pair queue, the first The second value in a pair of data. For the data pair queue, the first The first value of a pair of data.

[0036] For example, the first historical load data queue originates from load monitoring data during the past operation of the power system, which can be obtained through compliant methods such as power monitoring terminal collection and historical database retrieval. This data queue contains continuous load monitoring values, each corresponding to a specific collection timestamp, covering the load change trend within a specific time period. It provides basic data support for subsequent periodic inspections and stabilization processing. The continuity and integrity of the data are the core prerequisites for all subsequent processing steps.

[0037] The process involves periodic testing and differential processing of the first historical load data queue. This involves periodically testing the first historical load data queue and, based on the obtained first period length, performing differential processing to obtain the first stable queue. Since the first historical load data queue is typically affected by periodic factors (such as daily load fluctuations and weekly load fluctuations), exhibiting non-stationary characteristics, it cannot be directly applied to core business scenarios such as subsequent load forecasting. Therefore, it is necessary to clarify its inherent periodic patterns through periodic testing and then eliminate periodic fluctuations through differential processing to bring the data to a stable state, thereby improving the accuracy and reliability of subsequent data processing.

[0038] Figure 2 The diagram illustrates the principle of the first stable queue acquisition process provided by an embodiment of the present invention. The present invention performs periodic checks on the first historical load data queue 201 and performs differential processing on it based on the first period length obtained from the checks to obtain the first stable queue 203. The specific process is as follows: The first step is to determine the initial trial period (i.e., the length of the second period). The length of the second period serves as the initial trial parameter for periodic testing. Its initial value can be preset based on the regular fluctuation patterns of power load. For example, based on power system operating experience, it can be initially set to 24 hours (corresponding to the daily load cycle) or 168 hours (corresponding to the weekly load cycle). At the same time, the initial value can be flexibly adjusted according to the actual data collection frequency (such as collecting data once per hour or once every 15 minutes) to ensure that the initial period length has reasonable exploratory and practical characteristics.

[0039] The second step involves extracting data pairs 202 based on the second cycle length and constructing a data pair queue. The specific operation of the sliding extraction is as follows: starting from the beginning of the first historical load data queue 201, and sliding sequentially according to individual data collection units, with the second cycle length as the interval, a set of data pairs 202 (i.e., the load data at the current position and the load data at the corresponding position after the second cycle length interval) is extracted after each sliding operation, until no complete data pair 202 can be extracted, at which point extraction stops. All extracted valid data pairs 202 are then stored sequentially in the data pair queue according to the extraction order, ensuring that the data pairs 202 fully reflect the load correlation characteristics corresponding to the cycle at different time points.

[0040] The third step involves calculating the periodicity index based on the data in the queue and storing the result in the periodicity index queue. The periodicity index quantifies the degree of periodic correlation between the data and the data in the queue. The smaller the index value, the stronger the consistency of the fluctuation pattern between the two values ​​in the queue, and the closer the corresponding period length is to the actual period of the data. After calculation, the obtained periodicity index result is stored in the periodicity index queue, establishing a one-to-one correspondence with the corresponding second period length.

[0041] The fourth step is to determine whether the length of the second period has reached the preset threshold. The threshold needs to be determined comprehensively based on the total length of the first historical load data queue 201, the data collection frequency, and actual business needs. For example, it can be set to half the total length of the data queue to avoid insufficient data extraction for data pairs 202 due to an excessively long period, resulting in errors in the calculation of the periodic index, or the inability to reflect the true periodicity of the data due to an excessively short period. If the length of the second period has not reached the preset threshold, it is incremented or decremented sequentially (the increment / decrement range can be preset to a single collection unit or a fixed proportion), and the process returns to the step of "extracting data pairs 202 based on the second period length and constructing a data pair queue" to continue the periodicity check. If the length of the second period has reached the preset threshold, the periodicity testing process is terminated, and the optimal period length selection stage begins.

[0042] The fifth step is to select the optimal period length and determine the target data pair queue. The second period length corresponding to the minimum value in the periodic index queue is determined as the first period length (i.e., the true period of the first historical load data queue 201), and the queue of data pairs corresponding to this first period length is taken as the target data pair queue. Since the data pairs corresponding to the minimum value of the periodic index have the most consistent fluctuation patterns, their corresponding period lengths best reflect the inherent periodic characteristics of the data. The target data pair queue determined accordingly can provide a reliable basis for subsequent differential processing.

[0043] Step 6: Traverse and extract data pairs 202 from the target data pair queue. Following the storage order of the target data pair queue, traverse and extract all data pairs 202 one by one to ensure that no valid data is missed.

[0044] Step 7: Perform a difference operation on the extracted data pairs 202 to obtain the difference value. The difference operation is performed on each set of extracted data pairs 202 in a preset order, which can be set as "subtracting the previous value from the next value in data pair 202". This difference operation effectively eliminates periodic fluctuations in the data. The difference value corresponding to each set of data pairs 202 reflects the load fluctuation differences at corresponding time points in two cycles, thereby weakening the impact of periodic factors on data stability.

[0045] Step 8: Construct the first stabilization queue 203 based on the position of each data pair 202 in the target data pair queue. According to the position of each data pair 202 in the target data pair queue, all difference values ​​are arranged in the corresponding order to construct the first stabilization queue 203. After difference processing, the first stabilization queue 203 has eliminated the periodic fluctuations in the original load data, presenting a stable state, and can be directly applied to subsequent data processing stages such as load forecasting and anomaly detection.

[0046] In some implementations, the specific operation of calculating a periodic index on the queue based on data and storing the result in a periodic index queue includes: calculating a periodic index on the queue according to a first formula and data, and storing the calculated result in a periodic index queue, wherein the first formula is as follows:

[0047] The parameters in the formula are defined as follows: As a periodicity index, its value directly reflects the strength of the periodicity of the data in the queue. The smaller the PI value, the stronger the periodicity of the data. This refers to the total number of valid data pairs in the data pair queue, which is the number of complete data pairs extracted during this trial period after excluding invalid data such as missing values ​​and outliers. For the data pair queue, the first The second value of a data pair corresponds to the load data after the second cycle length in the data pair; For the data pair queue, the first The first value of each data pair corresponds to the load data at the initial position in the data pair. During the calculation of the periodic index, outlier data pairs in the data pair queue need to be preprocessed (e.g., missing value removal, outlier interpolation correction) to ensure the accuracy and representativeness of the calculation results; after the calculation is completed, the obtained values ​​are... The value is stored in the periodic index queue and associated with the corresponding second period length, providing reliable data support for the subsequent selection of the optimal period length.

[0048] In step 103, a load prediction equation is constructed based on the first stabilization queue, and data is extracted from the first stabilization queue and input into the load prediction equation to obtain the first load prediction data queue.

[0049] In some implementations, the step of constructing a load forecasting equation based on the first stabilization queue and extracting data from the first stabilization queue and inputting it into the load forecasting equation to obtain a first load forecasting data queue includes: Obtain the first fundamental time series equation and multiple first lengths; For each first length, multiple first data segments and multiple reference data are extracted from the first stabilization queue in a sliding manner according to the first length, and the extracted multiple first data segments and multiple reference data are constructed into a first dataset, wherein each first data segment corresponds to a reference data, and the reference data corresponding to the first data segment is connected to the first data segment in the first stabilization queue; For each first dataset, the number of parameters of the basic time series equation is adjusted according to the first length corresponding to the first dataset to obtain the second basic time series equation; Prediction bias acquisition steps: For each second basic time series equation, substitute multiple first data segments from the first dataset corresponding to the second basic time series equation into the second basic time series equation, and determine the prediction bias of the second basic time series equation based on the multiple outputs obtained and multiple reference data from the first dataset corresponding to the second basic time series equation. If there is a prediction deviation less than the deviation threshold among multiple prediction deviations, then the second basic time series equation corresponding to the smallest prediction deviation is taken as the load prediction equation, and the first length corresponding to the load prediction equation is taken as the equation data length. Otherwise, based on multiple prediction biases, the parameters of multiple second basic time series equations are adjusted, and the process jumps to the prediction bias acquisition step. Based on the length of the equation data, multiple second data segments are extracted from the first stabilization queue in a sliding manner; The multiple second data segments are input into the load prediction equation to obtain multiple differential prediction values; For each differential prediction value, the target historical load is extracted from the first historical load data queue based on the first cycle length and the differential prediction value; For each target historical load, the sum of the target historical load and the differential prediction value is added to the first load prediction data queue as load prediction data.

[0050] In some implementations, the first basic timing equation is:

[0051] In the formula, These are the difference prediction values. For the first The weight parameters for each input difference value. For the first Each input difference value For the intercept parameter, This is the total number of input difference values.

[0052] For example, the present invention constructs a load forecasting equation based on a first stabilization queue, extracts effective data from the first stabilization queue and inputs it into the load forecasting equation, and obtains a first load forecasting data queue through equation calculation. The first stabilization queue has eliminated the periodic fluctuations of the original load data and has stabilization characteristics. The load forecasting equation constructed based on this queue can effectively capture the inherent variation law of the load data and improve the accuracy of load forecasting.

[0053] The load forecasting equation is constructed based on the first stabilization queue, and data is extracted from the first stabilization queue and input into the load forecasting equation to obtain the first load forecasting data queue. The specific implementation process is as follows: The first step is to obtain the first basic time series equation and multiple first lengths. The first basic time series equation is the foundational model for load forecasting. It uses time series forecasting related equations, is adjustable, and can adapt to forecasting requirements with different data lengths. The multiple first lengths are time series window length parameters used to determine the length of the data segment input to the forecasting equation. Their values ​​need to be comprehensively set in combination with the length of the first stabilization queue, the data acquisition frequency, and the forecasting accuracy requirements. They can cover different time scales to ensure that the optimal equation input length is selected in subsequent screening.

[0054] The second step involves constructing the first dataset based on each first length. For each first length, a sliding extraction method is used to extract multiple first data segments and corresponding reference data from the first stabilization queue. All extracted first data segments and reference data are then constructed into the first dataset according to their correspondence. The length of each first data segment is equal to its corresponding first length. There is a clear connection between the reference data and the first data segment; specifically, the reference data is located immediately after the end of the corresponding first data segment in the first stabilization queue. This reference data is the first data point following the first data segment and is used to subsequently verify the accuracy of the prediction equation. During the sliding extraction process, the extraction proceeds sequentially by individual data collection units until it is impossible to extract a complete first data segment and its corresponding reference data, ensuring that the data samples in the first dataset are sufficiently representative.

[0055] The third step involves adjusting the number of parameters in the first basic time series equation based on the first length to obtain the second basic time series equation. For each first dataset, the number of parameters in the first basic time series equation is adjusted according to its corresponding first length to form a second basic time series equation adapted to that dataset. The adjustment of the number of parameters must match the first length; that is, the total number of input difference values ​​must be consistent with the length of the first data segment to ensure that the equation can fully utilize the effective information in the first data segment and improve the adaptability of the equation to the data.

[0056] The fourth step is to perform the prediction bias acquisition step to determine the prediction bias of each second basic time series equation. For each second basic time series equation, all the first data segments in the corresponding first dataset are substituted into the equation one by one, and multiple output values ​​are obtained through equation operation. The obtained output values ​​are compared with multiple reference data in the corresponding first dataset. The prediction bias of the second basic time series equation is quantitatively calculated using a preset bias calculation method (such as mean square error, mean absolute error, etc.). This bias directly reflects the prediction accuracy of the equation. The smaller the bias, the better the prediction effect.

[0057] The fifth step is to select the optimal load forecasting equation and its corresponding data length. The prediction deviations of all second basic time-series equations are assessed. If any of the prediction deviations is less than a preset deviation threshold, the second basic time-series equation with the smallest prediction deviation is selected as the final load forecasting equation. The first length of this equation is then determined as the equation data length. The deviation threshold needs to be set comprehensively based on actual forecasting accuracy requirements and the needs of the power system's business scenarios to determine whether the equation's prediction accuracy meets the standards. If multiple prediction deviations are not less than the preset deviation threshold, it indicates that the parameter settings of the current second basic time-series equations do not meet the forecasting requirements. In this case, the parameters (such as weight parameters and intercept parameters) of all second basic time-series equations need to be adjusted according to the magnitude and distribution of each prediction deviation. After adjustment, the process returns to the prediction deviation acquisition step described above, recalculating the prediction deviations of each equation until a load forecasting equation that meets the requirements is obtained.

[0058] Step 6: Extract the second data segment based on the equation data length using a sliding extraction method. According to the determined equation data length, extract multiple second data segments from the first stabilization queue using a sliding extraction method. The extraction method is consistent with the sliding extraction logic of the first data segment, that is, sliding sequentially according to a single data acquisition unit to extract continuous data segments with a length equal to the equation data length. This ensures that the extracted second data segments can meet the input requirements of the load forecasting equation, providing complete input data for subsequent forecasting calculations.

[0059] Step 7: Input the second data segment to obtain differential forecast values. Input the extracted second data segments one by one into the selected load forecast equation, and obtain multiple differential forecast values ​​through equation calculation. The differential forecast values ​​are load fluctuation forecast values ​​obtained based on stationary data, reflecting the load fluctuation trend at the corresponding time points, and need to be further restored to actual load forecast data.

[0060] Step 8: Extract the target historical load based on the first cycle length and the differential prediction value. For each differential prediction value, combined with the first cycle length determined in step 102, extract the corresponding target historical load from the first historical load data queue. The specific extraction logic is as follows: based on the position of the second data segment corresponding to the differential prediction value in the first stabilization queue, deduce its corresponding position in the first historical load data queue in reverse; combined with the first cycle length, extract the historical load data corresponding to that position as the target historical load, ensuring that the target historical load matches the time dimension of the differential prediction value.

[0061] The ninth step is to calculate the load forecast data and construct the first load forecast data queue. For each target historical load, sum it with the corresponding differential forecast value, and the result is the actual load forecast data. All load forecast data are added to the queue in their corresponding time order to form the first load forecast data queue. This queue reflects the load change forecast results for a future period of time and can be directly applied to core business operations such as power system dispatching and planning.

[0062] In some implementations, the first fundamental timing equation is as follows:

[0063] The parameters in the formula are defined as follows: The differential forecast value, i.e., the output of the load forecast equation, reflects the fluctuation forecast value of the stabilized load data; For the first The weight parameters for each input difference value are used to characterize the degree of influence of each input difference value on the predicted difference value. The initial value of the weight parameters can be preset based on experience and can be optimized through deviation adjustment in the future. For the first Each input difference value is a specific value in the first or second data segment of the input load prediction equation; The intercept parameter is used to correct the prediction bias of the equation and adapt to different load data scenarios. The total number of input difference values ​​is equal to the corresponding first length, i.e., the length of the input data segment, to ensure that the number of equation inputs matches the number of parameters.

[0064] In step 104, based on the photovoltaic expected output data queue and the first load prediction data queue, a first objective equation is constructed with the goal of minimizing abandoned energy, the volatility of photovoltaic output, and the charging and discharging capacity of the energy storage device, and constrained by the operating status of the energy storage device. The energy storage power is optimized based on the first objective equation to obtain the first energy storage power data queue.

[0065] In some implementations, the construction of a first objective equation based on the photovoltaic expected output data queue and the first load forecast data queue, with the objectives of minimizing energy curtailment, photovoltaic output volatility, and the charging and discharging capacity of the energy storage device, and constrained by the operating state of the energy storage device, includes: Based on the first load forecast data queue, a power balance equation is constructed to characterize the relationship between photovoltaic power output, energy storage device output, main power output, and load power. Based on the photovoltaic expected output data queue, construct the curtailment equation characterizing the relationship between photovoltaic output and curtailed energy, and the fluctuation equation characterizing the relationship between photovoltaic output and photovoltaic output volatility. Construct a charge / discharge equation to characterize the charge / discharge rate of an energy storage device; Different weighting coefficients are assigned to the curtailment of energy, the volatility of photovoltaic power output, and the charging and discharging capacity of energy storage devices, and a second objective equation is constructed based on the weighting coefficients; Construct constraint equations relating the state of charge (SBC) of an energy storage device to its charge and discharge limits; The first objective equation is obtained by combining the constraint equation with the second objective equation.

[0066] In some implementations, the first objective equation is:

[0067] In the formula, for Load power at time points, for Photovoltaic output at specific time points for Main power output at the time point The charging and discharging power of the energy storage device, To discard energy, This represents the total number of time points predicted for the first load forecast data queue. for Expected photovoltaic output at the specified time point The charge and discharge capacity of the energy storage device. This is the second objective equation. , as well as These are the first weighting coefficient, the second weighting coefficient, and the third weighting coefficient, respectively. for The state of charge of the energy storage device at a given time point. for The state of charge of the energy storage device at a given time point. The charging power conversion factor is... Rated reference power, This is the discharge smoothness coefficient. This is the charging smoothness coefficient.

[0068] For example, the core of the integrated source-grid-load-storage coordinated control method is to combine the photovoltaic expected output data queue with the first load forecast data queue obtained in the previous steps to construct a first objective equation for multi-objective optimization. The core optimization objectives are minimizing energy curtailment, minimizing photovoltaic output fluctuations, and minimizing the charging and discharging of energy storage devices. Simultaneously, the operating status of the energy storage devices is used as a constraint. By solving and optimizing the first objective equation, the optimal energy storage power at each time point is obtained, thereby constructing the first energy storage power data queue. This step is crucial for achieving coordinated scheduling of photovoltaic-energy storage-main power sources. Through multi-objective optimization, it can effectively improve photovoltaic energy utilization, smooth photovoltaic output fluctuations, extend the service life of energy storage devices, and ensure the safe, efficient, and economical operation of the power system.

[0069] Based on the photovoltaic expected output data queue and the first load forecast data queue, a first objective equation is constructed with the goals of minimizing energy curtailment, the volatility of photovoltaic output, and the charging and discharging capacity of energy storage devices, and with the operating status of energy storage devices as constraints. The specific implementation process is as follows: The first step is to construct a power balance equation. Using the load power at each time point in the first load forecast data queue as the core, and combining the principle of power supply and demand balance in the power system, a power balance equation is constructed to characterize the quantitative relationship between photovoltaic output, energy storage device output, main power output, and load power. This equation serves as the fundamental constraint for the entire objective optimization model, clarifying that within any given time point, the total system output (photovoltaic output + energy storage device output + main power output) must maintain a dynamic balance with the total system load power, ensuring stable power supply and avoiding system fluctuations caused by power shortages or excesses.

[0070] The second step involves constructing the curtailment energy equation and the photovoltaic (PV) output fluctuation equation. Based on the ideal PV output at each time point in the expected PV output data queue, and combined with the actual PV output at the corresponding time points in the first load forecast data queue, a curtailment energy equation is constructed to quantify the total amount of PV energy that is not fully utilized by the system, reflecting the utilization efficiency of PV energy. Simultaneously, based on the actual PV output at each time point in the first load forecast data queue, a PV output fluctuation equation is constructed to quantify the fluctuation amplitude of PV output between adjacent time points, providing a quantitative basis for smoothing PV fluctuations and ensuring the stability of the power system operation.

[0071] The third step is to construct the energy storage device's charge / discharge equation. By combining the charge and discharge power of the energy storage device at various time points, an equation characterizing the energy storage device's charge / discharge capacity is constructed. The charge / discharge capacity of the energy storage device is directly related to its operating losses and service life. Excessive charge / discharge will lead to frequent charging and discharging, accelerating equipment aging and increasing operating costs; insufficient charge / discharge will prevent it from fully utilizing its role in smoothing fluctuations and regulating power balance. Therefore, this equation is needed to quantify the charge / discharge capacity as one of the optimization objectives.

[0072] The fourth step involves assigning weight coefficients and constructing the second objective equation. Based on the actual operational needs of the power system (such as economic efficiency, stability, and equipment lifespan priority), specific weight coefficients (w1, w2, w3) are assigned to the three optimization objectives: energy curtailment, photovoltaic power output volatility, and energy storage device charging / discharging capacity. The weight coefficients all range from (0, 1) and satisfy w1 + w2 + w3 = 1, ensuring the rationality and balance of the weight allocation. Based on the assigned weight coefficients, the three optimization objectives are weighted and summed to construct the second objective equation. This transforms the multi-objective optimization problem into a single-objective optimization problem, simplifying the equation-solving process while also considering the priority requirements of each optimization objective.

[0073] The fifth step is to construct the operational constraint equations for the energy storage device. Combining the physical characteristics, safety operation requirements, and equipment parameters of the energy storage device, constraint equations are constructed to characterize the relationship between the state of charge (SOC) and charge / discharge limits. These constraint equations clarify the charge / discharge power boundaries of the energy storage device under different SOCs, preventing overcharging, over-discharging, and sudden changes in charge / discharge power. This ensures the safe and stable operation of the energy storage device and extends its service life, and is the core constraint condition of the first objective equation.

[0074] The sixth step involves solving the simultaneous equations to obtain the first objective equation. The power balance equation, the energy curtailment equation, the photovoltaic output fluctuation equation, the energy storage device charging / discharging equation, the second objective equation, and the energy storage device operation constraint equations constructed above are simultaneously integrated to form the complete first objective equation. This equation simultaneously encompasses multi-objective optimization requirements and equipment operation constraints, constituting a solvable multi-objective optimization model, providing the core theoretical basis for subsequent optimization of energy storage power.

[0075] In some implementations, the specific expression of the first objective equation, the detailed explanation of each parameter, and the calculation logic are as follows. The equation adopts a multi-constraint, multi-objective integrated form to comprehensively cover the optimization objective and operational constraints:

[0076] The detailed explanations of each equation and parameter are as follows: Power balance equation ( This equation is the core expression for the power supply and demand balance of the power system, and it clarifies the power distribution relationship at each time point.

[0077] in, for The load power at each time point is taken from the first load forecast data queue. The load forecast value at a given time point represents the total load demand of the system at that moment. for The photovoltaic output at each time point is taken from the first load forecast data queue. The actual photovoltaic output value at a given time point represents the actual power generation capacity of the photovoltaic system at that moment. for The main power output at a given time point represents the output power of the main power source (such as thermal power, hydropower, grid supplementary power, etc.) at that moment, and is used to make up for the gap between photovoltaic power output and energy storage device output to ensure power supply to the load; for The charging and discharging power of the energy storage device at a given time point is positive if the device is discharging and outputting power to the system, and negative if the device is charging and absorbing power from the system.

[0078] Discard energy equation ( ): Used to quantify the total amount of photovoltaic energy discarded within a forecast period, reflecting the utilization efficiency of photovoltaic energy.

[0079] in, To predict the total energy wastage during the period, the unit is consistent with the power unit (in conjunction with the time dimension). This represents the total number of forecast time nodes in the first load forecast data queue, i.e., the number of time nodes corresponding to the total duration of load forecasting. for The expected photovoltaic output at a given time point is taken from the photovoltaic expected output data queue. The ideal photovoltaic output value at a given time point represents the power generation of a photovoltaic system under ideal illumination and environmental conditions. The function to set negative values ​​to zero, when When the difference is the amount of photovoltaic energy wasted at that time point, that is, the portion of photovoltaic output exceeding the current system demand that is not utilized and is wasted; when When the difference is negative, it indicates that the photovoltaic output has not reached the ideal value and needs to be supplemented by other power sources. This difference is not included in the curtailed energy. The total curtailed energy... This is the sum of all positive differences.

[0080] Photovoltaic power output fluctuation equation ( ): Used to quantify the overall fluctuation range of photovoltaic power output within the prediction period. The smaller the fluctuation, the more conducive it is to the stable dispatch of the power system.

[0081] in, This is an indicator of photovoltaic power output volatility; the larger the value, the more drastic the fluctuation in photovoltaic power output. It is calculated by summing the squares of the differences in photovoltaic power output between two adjacent time points and then dividing by the number of time intervals. This eliminates the randomness of fluctuations at a single point in time, more accurately reflects the overall fluctuation characteristics of photovoltaic power output, and provides a quantitative basis for the optimization goal of smoothing fluctuations.

[0082] Energy storage device charge / discharge equation ( This is used to quantify the overall charge and discharge intensity of the energy storage device within a prediction period. The smaller the charge and discharge amount, the smoother the operation of the energy storage device and the lower the equipment loss. The charging and discharging capacity index of the energy storage device is calculated by the average of the squares of the charging and discharging power at each time point, which can effectively avoid the mutual cancellation of positive and negative power and accurately reflect the charging and discharging operation intensity of the energy storage device. The meaning of positive and negative values ​​is consistent with that in the power balance equation. After squaring, the intensity of charging and discharging can be uniformly characterized. After accumulating, the average value is taken to obtain the average charging and discharging intensity over the entire prediction period.

[0083] Second objective equation ( This equation is the core of multi-objective optimization. It unifies and quantifies the three optimization objectives of abandoned energy, photovoltaic power output fluctuation, and energy storage device charging and discharging into a single objective function, which is used to solve for the optimal solution.

[0084] in, This is the value of the second objective function; the smaller this value, the better the achievement of the three optimization objectives. , , These are the weighting coefficients for abandoned energy, photovoltaic power output volatility, and energy storage device charging and discharging capacity, respectively. The values ​​of these weighting coefficients need to be determined based on the actual needs of the power system; for example, if economic efficiency is emphasized, the coefficients should be increased. Focusing on system stability increases Focusing on increasing the lifespan of energy storage devices And must meet This ensures the reasonable allocation of weights.

[0085] Energy storage device charge state equation ( ): Used to calculate the real-time state of charge of energy storage devices at each time point, reflecting the remaining power level of the energy storage devices, and is the core basis for constraining the charging and discharging power of energy storage.

[0086] in, for The state of charge of the energy storage device at a given time point ranges from (0,1). This indicates that the energy storage device is fully charged. This indicates that the energy storage device is out of power; exceeding this range will cause damage to the equipment. for The state of charge of the energy storage device at the time node (predicted initial moment) is the initial parameter for optimization calculation and can be set according to the actual operating state of the energy storage device. The charging power conversion factor is determined by equipment parameters such as the charging efficiency and rated capacity of the energy storage device, and characterizes the conversion ratio between charging and discharging power and changes in state of charge. The cumulative charge and discharge power at all times before time point t, during charging. When the value is negative, the cumulative value is negative. Subtracting this cumulative value (i.e., adding the absolute value) increases the state of charge; during discharge... If the value is positive, the cumulative value is positive; subtracting this cumulative value results in a decrease in the state of charge.

[0087] Energy storage device charge and discharge constraint equations: This set of constraint equations clarifies the charge and discharge power boundaries of the energy storage device under different states of charge, ensuring the safe and stable operation of the equipment. Specifically, it includes two core constraints: Discharge constraint: .in, This is the rated reference power of the energy storage device, characterizing its maximum charge and discharge capability. The discharge smoothness coefficient is used to limit the rate of change of discharge power and avoid sudden changes in discharge power from impacting the equipment. The larger the coefficient, the smoother the change in discharge power. This constraint indicates that the discharge power of the energy storage device is limited by the current state of charge. The higher the state of charge, the greater the maximum discharge power, and the lower the state of charge, the smaller the maximum discharge power. At the same time, the discharge power must be greater than or equal to 0, thus defining the range of power values ​​for the discharge state.

[0088] Charging constraints: .in, This is the charging smoothness coefficient, used to limit the rate of change of charging power and avoid sudden changes in charging power. The larger the coefficient, the smoother the change in charging power. This constraint indicates that the charging power (negative value) of the energy storage device is limited by the current state of charge; the lower the state of charge, the larger the absolute value of the maximum charging power (i.e., ...). The more negative the charge level, the higher the absolute value of the maximum charging power, and the lower the absolute value of the maximum charging power. At the same time, the charging power must be less than 0. The range of power values ​​for the charging state should be clearly defined to avoid overcharging.

[0089] By solving the first objective equation in the above simultaneous equations, the optimal energy storage charging and discharging power at each time point can be obtained. Optimal for all time points Arranged in chronological order, the first energy storage power data queue can be constructed, providing a precise basis for energy storage power control for the coordinated scheduling of photovoltaic-energy storage-main power sources.

[0090] The present invention provides an implementation method for integrated source-grid-load-storage coordinated control, which first acquires a first historical load data queue; then, periodically checks the first historical load data queue and differentiates it based on the first period length obtained from the checks to obtain a first stabilization queue; next, a load prediction equation is constructed based on the first stabilization queue, and data extracted from the first stabilization queue is input into the load prediction equation to obtain a first load prediction data queue; finally, based on the photovoltaic expected output data queue and the first load prediction data queue, a first objective equation is constructed with the goal of minimizing energy curtailment, photovoltaic output volatility, and the charging and discharging capacity of the energy storage device, constrained by the operating state of the energy storage device; and the energy storage power is optimized based on the first objective equation to obtain a first energy storage power data queue.

[0091] The method of this invention constitutes a complete integrated collaborative control process for power generation, grid, load, and storage: Optimizing the forecasting model improves the accuracy and applicability of load forecasting. The true period of the first historical load data queue is determined through periodic verification, and then a first stable queue is constructed through differential processing. This effectively eliminates periodic fluctuations (such as daily and weekly load fluctuations) in the original load data, bringing the data to a stable state. This processing method solves the technical pain point that non-stationary data cannot be directly used for forecasting. Compared with directly using raw non-stationary data for forecasting, it significantly improves the accuracy of load forecasting, ensuring that the first load forecasting data queue can truly reflect future load changes, providing accurate load reference for subsequent energy storage optimization and system scheduling. By constructing datasets with multiple sets of first-length data, adjusting the time-series equation parameters, and selecting the optimal load forecasting equation, precise adaptation between the forecasting model and data characteristics is achieved. This step effectively avoids the problem of insufficient adaptability of a single model by iteratively optimizing deviations and adjusting model parameters. It enables the load forecasting equation to be dynamically adjusted according to data characteristics, which not only improves the forecasting accuracy but also enhances the adaptability of the forecasting process. It can adapt to scenarios with different time scales and different load fluctuation characteristics, providing accurate load forecasting support for subsequent photovoltaic-energy storage-main power coordinated scheduling and reducing scheduling errors caused by forecasting deviations.

[0092] Multi-objective collaborative optimization improves energy utilization efficiency and economy. With minimizing energy curtailment, photovoltaic power output volatility, and energy storage device charging / discharging as core objectives, a first objective equation is constructed in conjunction with energy storage operation constraints, and energy storage power is optimized, achieving a balanced improvement across multiple objectives. On the one hand, minimizing energy curtailment effectively improves the utilization efficiency of photovoltaic energy, reduces photovoltaic energy waste, and lowers power system energy losses, meeting the development needs of energy conservation and emission reduction. On the other hand, minimizing photovoltaic power output volatility mitigates the impact of photovoltaic power output fluctuations on the power system, reduces the regulation pressure on the main power source, lowers the operating costs of the main power source, and improves the economic efficiency of system operation.

[0093] This invention forms a complete closed loop of "data acquisition - data stabilization - load forecasting - energy storage optimization," with each step closely linked and mutually supportive, achieving intelligent control of the entire process from basic data processing to terminal scheduling optimization. Compared to fragmented processing flows, this closed-loop process enables coordinated optimization of each link, improving the precision and intelligence of power system scheduling. It can better adapt to the access needs of new energy sources such as photovoltaics, providing a feasible technical path for the efficient operation of new power systems and helping to achieve the dual goals of new energy consumption and stable system operation.

[0094] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0095] The following are embodiments of the apparatus of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0096] Figure 3 This is a functional block diagram of the integrated source-grid-load-storage coordinated control device provided in the embodiments of the present invention, with reference to... Figure 3 The integrated source-grid-load-storage coordinated control device includes: a historical load acquisition module 301, a load data processing module 302, a load forecasting module 303, and a comprehensive control module 304, wherein: Historical load acquisition module 301 is used to acquire the first historical load data queue; The load data processing module 302 is used to periodically check the first historical load data queue and perform differential processing on the first historical load data queue according to the first period length obtained by the check to obtain a first stable queue. The load forecasting module 303 is used to construct a load forecasting equation based on the first stabilization queue, and extract data from the first stabilization queue and input it into the load forecasting equation to obtain a first load forecasting data queue. The integrated control module 304 is used to construct a first objective equation based on the photovoltaic expected output data queue and the first load prediction data queue, with the goal of minimizing the curtailment of energy, the volatility of photovoltaic output and the charging and discharging of the energy storage device, and constrained by the operating status of the energy storage device. The module optimizes the energy storage power according to the first objective equation and obtains the first energy storage power data queue.

[0097] Figure 4 This is a functional block diagram of the electronic device provided in an embodiment of the present invention. For example... Figure 4 As shown, the electronic device 4 in this embodiment includes a processor 400 and a memory 401, wherein the memory 401 stores a computer program 402 that can run on the processor 400. When the processor 400 executes the computer program 402, it implements the steps of the above-described integrated source-grid-load-storage coordinated control method and embodiments, for example... Figure 1 Steps 101 to 104 are shown.

[0098] For example, the computer program 402 may be divided into one or more modules / units, which are stored in the memory 401 and executed by the processor 400 to complete the present invention.

[0099] The electronic device 4 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. The electronic device 4 may include, but is not limited to, a processor 400 and a memory 401. Those skilled in the art will understand that... Figure 4 This is merely an example of electronic device 4 and does not constitute a limitation on electronic device 4. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 4 may also include input / output devices, network access devices, buses, etc.

[0100] The processor 400 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0101] The memory 401 can be an internal storage unit of the electronic device 4, such as a hard disk or memory. The memory 401 can also be an external storage device of the electronic device 4, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 401 can include both internal and external storage units of the electronic device 4. The memory 401 is used to store the computer program 402 and other programs and data required by the electronic device 4. The memory 401 can also be used to temporarily store data that has been output or will be output.

[0102] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the aforementioned method embodiments, and will not be repeated here.

[0103] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0104] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0105] In the embodiments provided by this invention, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0106] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0107] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0108] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above-described embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various methods and apparatus embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0109] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for integrated source-grid-load-storage coordinated control, characterized in that, include: Obtain the first historical load data queue; The first historical load data queue is periodically tested, and the first historical load data queue is differentiated based on the first period length obtained from the test to obtain the first stable queue. A load forecasting equation is constructed based on the first stabilization queue, and data is extracted from the first stabilization queue and input into the load forecasting equation to obtain the first load forecasting data queue. Based on the photovoltaic expected output data queue and the first load prediction data queue, a first objective equation is constructed with the goal of minimizing abandoned energy, the volatility of photovoltaic output, and the charging and discharging capacity of energy storage devices, and constrained by the operating status of energy storage devices. The energy storage power is optimized based on the first objective equation to obtain the first energy storage power data queue.

2. The integrated source-grid-load-storage coordinated control method according to claim 1, characterized in that, The step of periodically inspecting the first historical load data queue and differentially processing the first historical load data queue based on the first period length obtained from the inspection to obtain a first stable queue includes: Obtain the length of the second cycle; Based on the second cycle length, data pairs are extracted from the first historical load data queue in a sliding manner, and the extracted data pairs are added to the data pair queue. The periodic index is calculated for the queue based on the data, and the calculation result is added to the periodic index queue. If the second cycle length does not reach the threshold, the second cycle length is incremented or decremented sequentially, and the process jumps to the step of extracting data pairs from the first historical load data queue in a sliding manner according to the second cycle length, and adding the extracted data pairs to the data pair queue. Otherwise, the second period length corresponding to the minimum value in the periodic index queue is taken as the first period length, and the data pair queue corresponding to the first period length is taken as the target data pair queue. Extract data pairs sequentially from the target data pair queue; For each extracted data pair, perform a difference operation in a predetermined order to obtain the difference value; Based on the position of the data pair in the target data queue, multiple difference values ​​are constructed into the first stabilization queue.

3. The integrated source-grid-load-storage coordinated control method according to claim 2, characterized in that, The step of calculating the periodic index of the queue based on the data and adding the calculation result to the periodic index queue includes: The periodic index is calculated for the queue based on the first formula and the data, and the calculated result is added to the periodic index queue. The first formula is: In the formula, It is a cyclical index. This represents the total number of data pairs in the data pair queue. For the data pair queue, the first The second value in a pair of data. For the data pair queue, the first The first value of a pair of data.

4. The integrated source-grid-load-storage coordinated control method according to claim 1, characterized in that, The step of constructing a load forecasting equation based on the first stabilization queue and extracting data from the first stabilization queue and inputting it into the load forecasting equation to obtain a first load forecasting data queue includes: Obtain the first fundamental time series equation and multiple first lengths; For each first length, multiple first data segments and multiple reference data are extracted from the first stabilization queue in a sliding manner according to the first length, and the extracted multiple first data segments and multiple reference data are constructed into a first dataset, wherein each first data segment corresponds to a reference data, and the reference data corresponding to the first data segment is connected to the first data segment in the first stabilization queue; For each first dataset, the number of parameters of the basic time series equation is adjusted according to the first length corresponding to the first dataset to obtain the second basic time series equation; Prediction bias acquisition steps: For each second basic time series equation, substitute multiple first data segments from the first dataset corresponding to the second basic time series equation into the second basic time series equation, and determine the prediction bias of the second basic time series equation based on the multiple outputs obtained and multiple reference data from the first dataset corresponding to the second basic time series equation. If there is a prediction deviation less than the deviation threshold among multiple prediction deviations, then the second basic time series equation corresponding to the smallest prediction deviation is taken as the load prediction equation, and the first length corresponding to the load prediction equation is taken as the equation data length. Otherwise, based on multiple prediction biases, the parameters of multiple second basic time series equations are adjusted, and the process jumps to the prediction bias acquisition step. Based on the length of the equation data, multiple second data segments are extracted from the first stabilization queue in a sliding manner; The multiple second data segments are input into the load prediction equation to obtain multiple differential prediction values; For each differential prediction value, the target historical load is extracted from the first historical load data queue based on the first cycle length and the differential prediction value; For each target historical load, the sum of the target historical load and the differential prediction value is added to the first load prediction data queue as load prediction data.

5. The integrated source-grid-load-storage coordinated control method according to claim 4, characterized in that, The first basic time series equation is: In the formula, These are the difference prediction values. For the first The weight parameters for each input difference value. For the first Each input difference value For the intercept parameter, This is the total number of input difference values.

6. The integrated source-grid-load-storage coordinated control method according to any one of claims 1-5, characterized in that, The first objective equation, constructed based on the photovoltaic expected output data queue and the first load prediction data queue, with the objectives of minimizing energy curtailment, photovoltaic output volatility, and the charging and discharging capacity of energy storage devices, and constrained by the operating state of energy storage devices, includes: Based on the first load forecast data queue, a power balance equation is constructed to characterize the relationship between photovoltaic power output, energy storage device output, main power output, and load power. Based on the photovoltaic expected output data queue, construct the curtailment equation characterizing the relationship between photovoltaic output and curtailed energy, and the fluctuation equation characterizing the relationship between photovoltaic output and photovoltaic output volatility. Construct a charge / discharge equation to characterize the charge / discharge rate of an energy storage device; Different weighting coefficients are assigned to the curtailment of energy, the volatility of photovoltaic power output, and the charging and discharging capacity of energy storage devices, and a second objective equation is constructed based on the weighting coefficients; Construct constraint equations relating the state of charge (SBC) of an energy storage device to its charge and discharge limits; The first objective equation is obtained by combining the constraint equation with the second objective equation.

7. The integrated source-grid-load-storage coordinated control method according to claim 6, characterized in that, The first objective equation is: In the formula, for Load power at time points, for Photovoltaic output at specific time points for Main power output at the time point The charging and discharging power of the energy storage device, To discard energy, This represents the total number of time nodes predicted for the first load forecast data queue. for The expected output of photovoltaic power at the specified time point. The charge and discharge capacity of the energy storage device. This is the second objective equation. , as well as These are the first weighting coefficient, the second weighting coefficient, and the third weighting coefficient, respectively. for The state of charge of the energy storage device at a given time point. for The state of charge of the energy storage device at a given time point. The charging power conversion factor. Rated reference power, This is the discharge smoothness coefficient. This is the charging smoothness coefficient. Function to set negative values ​​to zero.

8. A source-grid-load-storage integrated collaborative control device, characterized in that, For implementing the integrated source-grid-load-storage coordinated control method as described in any one of claims 1-7, the integrated source-grid-load-storage coordinated control device comprises: The historical load acquisition module is used to acquire the first historical load data queue; The load data processing module is used to periodically check the first historical load data queue and perform differential processing on the first historical load data queue according to the first period length obtained by the check to obtain a first stable queue. The load forecasting module is used to construct a load forecasting equation based on the first stabilization queue, and extract data from the first stabilization queue and input it into the load forecasting equation to obtain a first load forecasting data queue. as well as, The integrated control module is used to construct a first objective equation based on the photovoltaic expected output data queue and the first load prediction data queue, with the goal of minimizing the curtailment of energy, the volatility of photovoltaic output and the charging and discharging of energy storage devices, and constrained by the operating status of energy storage devices. The module optimizes the energy storage power based on the first objective equation and obtains the first energy storage power data queue.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7 above.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7 above.

Images