Micro-grid stability control method and system applied to new energy charging station
By dynamically adjusting the smoothing parameters of the time series forecasting algorithm and combining the changing patterns and abrupt changes in load data, the problem of low accuracy in load data forecasting for microgrids in new energy charging stations has been solved, thereby improving the stability of the microgrid and the safety of the equipment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ECONOMIC TECH RES INST OF STATE GRID HENAN ELECTRIC POWER
- Filing Date
- 2025-12-10
- Publication Date
- 2026-07-21
AI Technical Summary
In existing technologies, the accuracy of microgrid load data prediction in new energy charging stations is low, which leads to a decrease in microgrid stability and even damage to electrical equipment.
By dynamically adjusting the smoothing parameters of the trend component in the time series forecasting algorithm, and combining the approximation of changes in load data, the regularity of trends, and the degree of abrupt changes, appropriate forecasting parameters are dynamically selected to improve the accuracy of load data forecasting.
It improves the accuracy of microgrid load data prediction, enhances the stability of microgrids, and avoids equipment damage caused by prediction errors.
Smart Images
Figure CN121395502B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of power grid control technology, specifically to a microgrid stability control method and system applied to new energy charging stations. Background Technology
[0002] A microgrid is a small-scale power system consisting of distributed power sources, energy storage devices, loads, and corresponding control and protection devices. It can operate in parallel with an external power grid or independently. With the increasing popularity of new energy vehicles, the demand for new energy charging stations is growing, and the load share of charging stations in microgrids is also gradually increasing. The power consumption of charging piles largely determines the stability of the microgrid. Therefore, it is necessary to control and regulate the microgrid to ensure its stable operation.
[0003] Microgrids have limited output power and need to be connected to the main grid when the load is high. The main grid then provides power resources to compensate for the increased output power of the microgrid. Typically, if the microgrid's output power exceeds the load's power consumption, it can operate independently in islanded mode (off-grid mode). Conversely, if the output power is less than the load's power consumption, it needs to be connected to the main grid to ensure power output. Currently, time series forecasting algorithms are commonly used to predict load data for future periods to control the microgrid's connection to and from the grid.
[0004] However, the load data of the microgrid connected to the new energy charging station is easily affected by the usage of the charging station. As a result, the time series prediction algorithm uses a fixed smoothing parameter for the trend component, which makes it difficult to accurately predict the trend changes of the microgrid load data. This leads to low accuracy of the prediction results, which in turn affects the control of the microgrid, ultimately causing a decline in the stability of the microgrid and even damage to the electrical equipment. Summary of the Invention
[0005] In view of the above, it is necessary to provide a microgrid stability control method and system for new energy charging stations. Compared with traditional microgrid stability control methods for new energy charging stations, this method improves the accuracy of microgrid load data prediction by dynamically adjusting the smoothing parameter of the trend component in the time series prediction algorithm, thereby improving the control precision of the microgrid and enhancing its stability. In a first aspect, embodiments of this application provide a microgrid stability control method applied to new energy charging stations, the method comprising the following steps: Obtain the load data of the microgrid at each collection time before the current time within the same day, and the load data at each collection time within the previous day; The load data for the current day and the previous day are obtained in time series. Based on the similarity of the load data of the same number of peaks between the current day and the previous day, as well as the difference in the time of occurrence and the difference in the duration of peaks between the same number of peaks, the approximation of the change at the current moment is obtained. Combined with the difference in the frequency of occurrence of peaks between the current day and the previous day, the regularity of the load trend at the current moment is obtained. The current day and the previous day are divided into intervals of preset lengths and numbered in reverse order of time. The load mutation rate at the current moment is obtained by combining the difference in the dispersion of load data between different intervals before the current moment within the same day, the difference in the average level of load data within intervals with the same sequence number between the current day and the previous day, and the duration of the peaks of each peak point before the current moment within the same day. The parameter adjustment coefficient for the current moment is obtained by combining the load abrupt change rate and the load trend regularity; the trend smoothing coefficient for the current moment is obtained by combining the parameter adjustment coefficient with the preset initial trend smoothing coefficient; the load data for a preset period after the current moment is predicted by combining the trend smoothing coefficient with the time series prediction algorithm; and the grid connection and disconnection control of the microgrid is performed based on the prediction results.
[0006] In one embodiment, the process of changing the approximation is as follows: The load data between the adjacent previous trough and adjacent subsequent trough of each peak point are used to form the load data sequence of each peak point, and the fitting curve of each load data sequence is obtained; the similarity of the fitting curves corresponding to the peak points with the same sequence number between the current day and the previous day is calculated. The data collection times for the current day and the previous day are numbered sequentially, and the difference in the sequence number of the collection times of the peak points with the same sequence number between the current day and the previous day is calculated. The time interval between the adjacent preceding trough and the adjacent following trough of each peak is counted, and the difference in the time interval corresponding to the peak with the same sequence number between the current day and the previous day is calculated. Calculate the sum of the similarities, the cumulative sum of the differences, and the cumulative sum of the differences for all peak points with the same serial number between the current day and the previous day; The approximation of the change can be further obtained through the sum, the accumulated value, and the accumulated sum.
[0007] In one embodiment, the method for calculating the approximation of the change is as follows: calculate the product of the accumulated value and the accumulated sum, wherein the approximation of the change is the ratio of the sum to the product.
[0008] In one embodiment, the process of obtaining the load trend regularity is as follows: The frequency of load data occurring before the current time within the same day is recorded as the first frequency; the frequency of load data occurring before the time with the same sequence number as the current time within the previous day is recorded as the second frequency; the deviation between the first frequency and the second frequency is calculated; the cumulative value of the deviation and a preset positive number is calculated. The regularity of the load trend is the ratio of the approximation of the change to the cumulative value.
[0009] In one embodiment, the formula for calculating the load mutation degree is: In the formula, This represents the load mutation degree at the i-th data collection time within the day; , These represent the average load data within the first interval before the i-th collection time on the same day and the previous day, respectively. , represents the dispersion of the load data in the 1st and yth intervals before the i-th collection time within the day, respectively; Y represents the number of intervals at the i-th collection time within the day; It represents the average of the time intervals corresponding to all peak points in the first interval before the i-th collection time on the same day; This indicates the absolute value operation.
[0010] In one embodiment, the process of obtaining the parameter adjustment coefficient is as follows: mapping the load trend regularity to a first positive number, wherein the parameter adjustment coefficient is twice the normalized value of the ratio of the load mutation degree to the first positive number.
[0011] In one embodiment, the trend smoothing coefficient at the current moment is the product of the parameter adjustment coefficient and the preset initial trend smoothing coefficient.
[0012] In one embodiment, the method for predicting load data within a preset time period after the current moment is as follows: All load data collected before the current time within the day are arranged in time sequence to form a microgrid load data sequence. The microgrid load data sequence is used as the input of a time series prediction algorithm, and the output is the load data of a preset number of collection times after the current time. The trend smoothing parameter of the current time is used as the smoothing parameter of the trend component in the time series prediction algorithm.
[0013] In one embodiment, the step of controlling the microgrid for grid connection or disconnection based on the prediction results includes: Obtain the output power of the microgrid at the current moment. If the output power is greater than or equal to the load data of all collected moments after the current moment as predicted, switch the microgrid to island mode; otherwise, connect the microgrid to the main grid.
[0014] Secondly, embodiments of this application also provide a microgrid stability control system for new energy charging stations, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described microgrid stability control methods for new energy charging stations.
[0015] This application has at least the following beneficial effects: This application quantifies the similarity between current load data and historical data in terms of fluctuation trends by calculating the approximation of changes, which can provide a basis for subsequent adjustment of prediction parameters. Furthermore, by combining the similarity of load data fluctuation frequencies, the load trend regularity is calculated, which can assess the regularity and stability of load data changes. This is beneficial for dynamically selecting appropriate prediction parameters based on the changes in load data. When the regularity of load changes is stronger, a smoothing parameter with a smaller trend component is used to improve the stability of the prediction. Conversely, a smoothing parameter with a larger trend component is used to improve the sensitivity of the prediction. Furthermore, by calculating the load abruptness, sudden changes in the load data can be effectively identified. When the load data is more likely to undergo abrupt changes, the smoothing coefficient of the trend component during prediction should be increased. This helps to improve the sensitivity to changes in the trend of the load data and avoid prediction errors caused by abrupt changes in the load data. Furthermore, by dynamically adjusting the smoothing parameters of the trend component in the time series forecasting algorithm based on the changing patterns and abrupt changes in load data, the accuracy of microgrid load data prediction is improved. This addresses the current problem that when using time series forecasting algorithms to predict the load data of microgrids connected to charging stations, the load data is easily affected by the usage of the charging stations. Using fixed smoothing parameters for the trend component makes it difficult to accurately predict the trend changes in the microgrid load data, resulting in low prediction accuracy. This, in turn, affects the control of the microgrid, leading to decreased microgrid stability and damage to new energy vehicles. Attached Figure Description
[0016] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, 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 this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0017] Figure 1 A flowchart illustrating the steps of a microgrid stability control method for new energy charging stations, provided as an embodiment of this application; Figure 2This is a schematic diagram illustrating the process of obtaining the approximation of the change. Figure 3 This is a schematic diagram illustrating the process of obtaining the trend smoothing coefficient. Detailed Implementation
[0018] In the description of the embodiments in this application, the words "exemplary," "or," and "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary," "or," and "for example" is intended to present the relevant concepts in a specific manner.
[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. It should be understood that, unless otherwise stated, " / " in this application means "or".
[0020] It should also be noted that the terms "first" and "second" in this application are used to distinguish similar objects, rather than to describe a specific order or sequence.
[0021] The following description, in conjunction with the accompanying drawings, details the specific scheme of the microgrid stability control method and system provided in this application for application in new energy charging stations.
[0022] Please see Figure 1 The diagram illustrates a flowchart of a microgrid stability control method for new energy charging stations according to an embodiment of this application. The method includes the following steps: Step 1: Obtain the load data of the microgrid at each collection time before the current time within the same day, and the load data at each collection time within the previous day.
[0023] This tool acquires load data for the microgrid at each time point before the current time of the current day, as well as load data at each time point of the previous day. Load data refers to the electrical power consumed by all electrical devices in the microgrid, reflecting the power demand of the loads in the microgrid at each acquisition time.
[0024] In this embodiment, load data is collected every 5 minutes. The 5-minute interval is just one embodiment of this application. Implementers can set it according to the actual situation. This application does not impose any special restrictions.
[0025] Step 2: Obtain the peak points of the load data in the time series for the current day and the previous day respectively. According to the time series number, obtain the approximation of the change at the current moment by the similarity of the preset neighbor load data of the peak points with the same sequence number between the current day and the previous day, as well as the difference in the occurrence time and the difference in the duration of the peaks between the peak points with the same sequence number. Combined with the difference in the occurrence frequency of the peak points between the current day and the previous day, obtain the regularity of the load trend at the current moment.
[0026] For the microgrid connected to new energy charging stations, the load generated during each use of the charging station is similar. When a new energy vehicle is charging, it goes through two stages: constant current charging and constant voltage charging. Therefore, the load of a single charging station initially increases and then decreases. Although the usage of an individual charging station is affected by user behavior, there are certain regularities in overall charging station usage. Peak usage times are midday and from evening to early morning, with some usage in the morning and afternoon. Consequently, for the microgrid connected to the charging station, the peak load times are midday and from evening to early morning, with some fluctuations at other times. If the daily usage regularity of the charging stations is high, the trend of the microgrid's load data will also be similar. Therefore, a relatively small trend smoothing parameter can be selected during the prediction process to ensure the accuracy of the predicted trend. Specifically, by analyzing the load data of the current day and the previous day, it can be found that the peak and trough trends of the load data in the time series are quite similar to historical data, and the overall fluctuation frequency is also relatively consistent. However, the use of charging stations inevitably involves a large degree of randomness. For example, during off-peak hours, a sudden and rapid increase in load may occur due to a large number of vehicles connecting to the charging station at the same time, or the charging power of different types of vehicles may differ, causing the load data of the charging station to change significantly from historical data at certain times. This, in turn, leads to a significant change in the load data of the microgrid. If a low trend smoothing parameter is still used, the sensitivity to changes in load data trends may be reduced, resulting in the prediction results failing to accurately reflect the actual load change trend.
[0027] Based on the above analysis, the peak points of the load data in the time series for the current day and the previous day are obtained respectively. According to the time series number, the load data between the adjacent preceding trough point and the adjacent following trough point of each peak point are combined to form the load data sequence of each peak point, and the fitting curve of each load data sequence is obtained. According to the time series number, the difference value of the sequence number of the peak point with the same sequence number between the current day and the previous day is calculated. The time interval between the adjacent preceding trough point and the adjacent following trough point of each peak point is counted, and the difference in the time interval corresponding to the peak point with the same sequence number between the current day and the previous day is calculated.
[0028] In this embodiment, a peak and trough detection algorithm is used to obtain the peak and trough points of the load data in the time series for the current day and the previous day. The peak and trough detection algorithm is a well-known technology and will not be described in detail in this application. As other implementation methods, based on the ability to obtain the peak and trough points of the load data in the time series for the current day and the previous day, implementers may use other existing technologies, such as extreme point detection algorithms, etc. This application does not impose any special restrictions.
[0029] In this embodiment, the least squares method is used to obtain the fitting curve of each load data sequence. The least squares method is a well-known technique and will not be described in detail in this application. As other implementation methods, based on the ability to obtain the fitting curve of each load data sequence, the implementer may use other existing techniques, such as local weighted regression, K-nearest neighbor regression, etc. This application does not impose any special restrictions.
[0030] Furthermore, by considering the similarity of the fitted curves corresponding to the peak points with the same serial number between the current day and the previous day, as well as the difference value and the difference amount, the approximation of the change at the current moment is obtained, expressed as: In the formula, This indicates the approximation of the change at the i-th data collection time within the same day; This represents the similarity between the fitted curve of the m-th peak point before the i-th collection time on the same day and the m-th peak point on the previous day; This represents the difference value corresponding to the m-th peak point before the i-th collection time within the same day; This represents the difference corresponding to the m-th peak point before the i-th collection time within the day; M represents the minimum value between the number of peak points within the day and the number of peak points within the previous day.
[0031] In this embodiment, the least squares method is used to obtain the fitting curve. The least squares method is a well-known technique and will not be described in detail in this application. As other implementation methods, based on the ability to obtain the fitting curves of each load data sequence, the implementer may use other existing techniques, such as local weighted regression, K-nearest neighbor regression, etc. This application does not impose any special restrictions.
[0032] In this embodiment, the similarity between fitted curves is calculated as follows: the fitted values of the load data in the load data sequence on their fitted curves are arranged in time sequence to form a fitted sequence, and the reciprocal of the sum of the DTW (Dynamic Time Warping) distance between two fitted sequences and a preset positive number is calculated. The calculation of the DTW distance is a well-known technique and will not be elaborated upon in this application. The preset positive number is used to avoid a denominator of 0, and the value of the preset positive number is 0.01. The implementer can set this value as they see fit, and this application does not impose any special restrictions.
[0033] In this embodiment, the difference in the sequence number of the acquisition time is the absolute value of the difference. As for other implementation methods, based on the ability to measure the degree of difference in the sequence number of the acquisition time, the implementer may use other calculation methods, such as the square of the difference, etc. This application does not impose any special restrictions.
[0034] In this embodiment, the difference in time intervals is the absolute value of the difference. As for other implementation methods, based on the measurable degree of difference in time intervals, the implementer may use other calculation methods, such as the square of the difference, etc. This application does not impose any special restrictions.
[0035] It should be noted that the higher the similarity between the load data collected by the microgrid within a given day and historical data in terms of fluctuation trends, the closer the times of change, and the closer the duration of peaks in the load data, the more similar the changes in the load data within that day are to the changes in historical data, and the greater the calculated approximation of the change. A schematic diagram of the process for obtaining the approximation of the change is shown below. Figure 2 As shown.
[0036] Furthermore, by combining the approximation of changes at the current moment with the deviation of the frequency of peak occurrences between the current day and the previous day, the regularity of the load trend at the current moment is obtained, expressed as: ; This indicates the regularity of the load trend at the i-th data collection time within the day; This represents the approximation of the change at the i-th collection time within the day; α represents a preset positive number used to avoid the denominator being 0. The value of α is preset by the user and can be set by the implementer. In this embodiment, the value of α is 0.01. This indicates the frequency of load data occurrences before the i-th data collection time within the same day; This indicates the frequency of occurrence of load data before the time with the same sequence number as the i-th collection time within the previous day; This indicates the absolute value operation. Recorded as the first frequency; This is denoted as the second frequency.
[0037] It should be noted that: the greater the approximation of the change, the higher the degree of similarity between the load data collected on the same day and the historical data in terms of fluctuation trend, and the closer the time of change. At the same time, the duration of the peaks in the load data is also closer. If the approximation of the overall fluctuation frequency of the load data is higher up to the i-th collection time, then the degree of change of the microgrid load data is more similar to the degree of change of the historical data. Therefore, when predicting the load data, the smoothing parameter of the trend component should be smaller to ensure the stability of the trend prediction.
[0038] Step 3: Divide the current day and the previous day into intervals of preset lengths and number them in reverse order of time. By considering the difference in the dispersion of load data between different intervals before the current time within the same day, the difference in the average level of load data within intervals with the same sequence number between the current day and the previous day, and the duration of the peaks of each peak point before the current time within the same day, obtain the load mutation degree at the current time.
[0039] Due to variations in the coverage of different microgrids, some microgrids may only supply power to charging stations within cities, while others may cover both urban and highway service area charging stations. The number of charging stations in highway service areas is relatively small compared to those in cities, resulting in a smaller impact on the overall microgrid when loads are generated. Furthermore, the peak load patterns of highway service area charging stations are less predictable than those in cities. Consequently, when peak usage occurs at highway service area charging stations, the current load data for the microgrid still closely approximates historical data, leading to the use of lower trend component smoothing parameters. This further reduces the sensitivity to trend changes in load data prediction, lowering prediction accuracy and ultimately affecting the stability control precision of the microgrid. Therefore, further analysis is needed.
[0040] For charging stations in highway service areas, the types of vehicles using them are relatively complex, including pure electric vehicles, plug-in hybrid electric vehicles, and large pure electric vehicles of different sizes. In cities, however, pure electric vehicles are more prevalent. Different types of vehicles generate different loads during charging; for example, hybrid electric vehicles generate relatively less load during charging compared to pure electric vehicles. Therefore, when the load generated by charging stations in highway service areas during peak usage periods overlaps with the charging load of vehicles in the city, the microgrid load experiences short-term peak fluctuations. Furthermore, due to the overlapping loads, the microgrid load increases to some extent in the short term. Additionally, vehicles charging in highway service areas typically do so temporarily, resulting in high frequency but short duration of charging station usage. Specifically, at a given moment, the peak value of the microgrid load data shows short-term peak fluctuations, with an overall increase in the peak value. The frequency of load peaks is high in the short term, but the duration of these peaks is short.
[0041] Based on the above analysis, starting from the current time of the day, each consecutive preset number of collection times is divided into an interval and numbered in reverse chronological order. The closer the interval is to the current time, the smaller the number. At the same time, starting from the time with the same number as the current time in the previous day, each consecutive preset number of collection times is divided into an interval and numbered in reverse chronological order.
[0042] In this embodiment, the preset quantity is 30. The preset quantity is manually set, and the implementer can set it according to the actual situation. This application will not elaborate further. If there are cases where the data collection time is less than 30 during the division process, no processing will be performed.
[0043] Furthermore, by considering the dispersion difference of load data between different intervals before the current time within the same day, and the difference in the average level of load data within intervals with the same index between the current day and the previous day, combined with the peak duration of each peak point before the current time within the same day, the load abrupt change degree at the current time is obtained, expressed as: In the formula, This represents the load mutation degree at the i-th data collection time within the day; , These represent the average load data within the first interval before the i-th collection time on the same day and the previous day, respectively. , represents the dispersion of the load data in the 1st and yth intervals before the i-th collection time within the day, respectively; Y represents the number of intervals at the i-th collection time within the day; It represents the average of the time intervals corresponding to all peak points in the first interval before the i-th collection time on the same day; This indicates the absolute value operation.
[0044] In this embodiment, the dispersion is variance. As for other implementation methods, in addition to being able to measure the unevenness of the distribution of load data, implementers may use other existing technologies, such as standard deviation, coefficient of variation, etc., which will not be elaborated here.
[0045] It should be noted that: the greater the difference in the dispersion of load data between the first interval and the other intervals before the i-th data collection time, the more obvious the difference in the average load and the shorter the peak duration, the greater the fluctuation in load data in the short period before the i-th data collection time. This is more likely to be caused by the peak usage of charging stations in a certain highway service area. In this case, the possibility of sudden changes in microgrid load data is greater, and the sensitivity to load data trend changes should be increased, and the smoothing coefficient of the trend component during prediction should be increased.
[0046] Step 4: Obtain the parameter adjustment coefficient at the current moment by using the load abrupt change degree and the load trend regularity degree; obtain the trend smoothing coefficient at the current moment by using the parameter adjustment coefficient and a preset initial trend smoothing coefficient; predict the load data within a preset period after the current moment by using the trend smoothing coefficient and a time series prediction algorithm; and perform grid connection / disconnection control on the microgrid based on the prediction results.
[0047] The parameter adjustment coefficients for the current moment are obtained by combining the load abruptness and load trend regularity at the current moment. Specifically: The current load trend regularity is mapped to a first positive number. Twice the normalized value of the ratio of the current load abrupt change to the first positive number is used as the parameter adjustment coefficient for the current moment. Specifically, calculating twice the normalized value maps the normalized value to the range (0,2), thereby reducing or increasing the preset initial trend smoothing coefficient.
[0048] In this embodiment, the load trend regularity is mapped to a first positive number by calculating the sum of the load trend regularity and a preset positive number, wherein the preset positive number is 0.01.
[0049] In another embodiment, the load trend regularity is mapped to a first positive number by using the sum of the load trend regularity and a value greater than 1 as the argument of the logarithmic function, where the value greater than 1 is 1.01.
[0050] In this embodiment, the hyperbolic tangent function is used to obtain the normalized value of the ratio of the current load mutation degree to the first positive number. The hyperbolic tangent function is a well-known technology and will not be described in detail in this application.
[0051] It should be noted that the lower the trend regularity of the load data collected before the i-th acquisition time and the higher the load abruptness of the microgrid, the greater the trend change that the load data collected at the i-th acquisition time may undergo. In this case, a higher trend smoothing parameter should be used to improve the sensitivity of load data prediction to trends, thereby improving the accuracy of data prediction.
[0052] Furthermore, by combining the parameter adjustment coefficient at the current moment with the preset initial trend smoothing coefficient, the trend smoothing coefficient at the current moment is obtained, expressed as: In the formula, This represents the trend smoothing coefficient at the i-th data collection time within the day; This indicates the preset initial trend smoothing coefficient; This represents the parameter adjustment coefficient at the i-th data collection time within the day. A schematic diagram illustrating the process of obtaining the trend smoothing coefficient is shown below. Figure 3 As shown.
[0053] In this embodiment, the initial trend smoothing coefficient is preset to 0.5. The initial trend smoothing coefficient is preset by the user and can be set by the implementer according to the actual situation. When the trend smoothing coefficient is greater than 1, the trend smoothing coefficient is set to 1.
[0054] Furthermore, by using the trend smoothing coefficient at the current moment and combining it with a time series forecasting algorithm, the load data for a preset period after the current moment is predicted, specifically as follows: All load data collected before the current time within the day are arranged in time sequence to form a microgrid load data sequence. The microgrid load data sequence is used as the input of a time series prediction algorithm, and the output is the load data of a preset number of collection times after the current time. The trend smoothing parameter of the current time is used as the smoothing parameter of the trend component in the time series prediction algorithm.
[0055] In this embodiment, the time series prediction algorithm is the Error-Trend-Seasonality model. The prediction step size in the Error-Trend-Seasonality model is set to 10, which means that the load data of the 10 collection times after the current time is output. The value of the prediction step size is preset by the user. The implementer can set it according to the actual situation. This application does not impose any special restrictions. The Error-Trend-Seasonality model is a well-known technology and will not be described in detail here.
[0056] Obtain the output power of the microgrid at the current moment. If the output power is greater than or equal to the load data of all collected moments after the current moment as predicted, switch the microgrid to island mode; otherwise, connect the microgrid to the main grid.
[0057] Based on the same inventive concept as the above methods, this application also provides a microgrid stability control system for new energy charging stations, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described microgrid stability control methods for new energy charging stations.
[0058] In summary, this application quantifies the similarity between current load data and historical data in terms of fluctuation trends by calculating the approximation of changes, which can provide a basis for subsequent adjustment of prediction parameters. Furthermore, by combining the similarity of load data fluctuation frequencies, the load trend regularity is calculated, which can assess the regularity and stability of load data changes. This is beneficial for dynamically selecting appropriate prediction parameters based on the changes in load data. When the regularity of load changes is stronger, a smoothing parameter with a smaller trend component is used to improve the stability of the prediction; conversely, a smoothing parameter with a larger trend component is used to improve the sensitivity of the prediction. Furthermore, by calculating the load abruptness, sudden changes in the load data can be effectively identified. When the load data is more likely to undergo abrupt changes, the smoothing coefficient of the trend component during prediction should be increased. This helps to improve the sensitivity to changes in the trend of the load data and avoid prediction errors caused by abrupt changes in the load data. Furthermore, by dynamically adjusting the smoothing parameters of the trend component in the time series forecasting algorithm based on the changing patterns and abrupt changes in load data, the accuracy of microgrid load data prediction is improved. This addresses the current problem that when using time series forecasting algorithms to predict the load data of microgrids connected to charging stations, the load data is easily affected by the usage of the charging stations. Using fixed smoothing parameters for the trend component makes it difficult to accurately predict the trend changes in the microgrid load data, resulting in low prediction accuracy. This, in turn, affects the control of the microgrid, leading to decreased microgrid stability and damage to new energy vehicles.
[0059] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. In the descriptions corresponding to the flowcharts and block diagrams in the accompanying drawings, the operations or steps corresponding to different blocks may also occur in a different order than disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. Each block in a block diagram and / or flowchart, and combinations of blocks in a block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.
[0060] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from its essential characteristics. Therefore, the embodiments described above should be considered exemplary and non-limiting in all respects.
Claims
1. A microgrid stability control method applied to new energy charging stations, characterized in that, The method includes the following steps: Obtain the load data of the microgrid at each collection time before the current time within the same day, and the load data at each collection time within the previous day; The load data for the current day and the previous day are obtained in time series. Based on the similarity of the load data of the same number of peaks between the current day and the previous day, as well as the difference in the time of occurrence and the difference in the duration of peaks between the same number of peaks, the approximation of the change at the current moment is obtained. Combined with the difference in the frequency of occurrence of peaks between the current day and the previous day, the regularity of the load trend at the current moment is obtained. The current day and the previous day are divided into intervals of preset lengths and numbered in reverse order of time. The load mutation rate at the current moment is obtained by combining the difference in the dispersion of load data between different intervals before the current moment within the same day, the difference in the average level of load data within intervals with the same sequence number between the current day and the previous day, and the duration of the peaks of each peak point before the current moment within the same day. The parameter adjustment coefficient for the current moment is obtained by combining the load abrupt change rate and the load trend regularity; the trend smoothing coefficient for the current moment is obtained by combining the parameter adjustment coefficient with the preset initial trend smoothing coefficient; the load data for a preset period after the current moment is predicted by combining the trend smoothing coefficient with the time series prediction algorithm; and the grid connection and disconnection control of the microgrid is performed based on the prediction results. The process of determining the approximation of the change is as follows: The load data between the adjacent previous trough and adjacent subsequent trough of each peak point are used to form the load data sequence of each peak point, and the fitting curve of each load data sequence is obtained; the similarity of the fitting curves corresponding to the peak points with the same sequence number between the current day and the previous day is calculated. The data collection times for the current day and the previous day are numbered sequentially, and the difference in the sequence number of the collection times of the peak points with the same sequence number between the current day and the previous day is calculated. The time interval between the adjacent preceding trough and the adjacent following trough of each peak is counted, and the difference in the time interval corresponding to the peak with the same sequence number between the current day and the previous day is calculated. Calculate the sum of the similarities, the cumulative sum of the differences, and the cumulative sum of the differences for all peak points with the same serial number between the current day and the previous day; The approximation of the change can be further obtained through the sum, the accumulated value, and the accumulated sum; The process of obtaining the regularity of the load trend is as follows: The frequency of load data occurring before the current time within the same day is recorded as the first frequency; the frequency of load data occurring before the time with the same sequence number as the current time within the previous day is recorded as the second frequency; the deviation between the first frequency and the second frequency is calculated; the cumulative value of the deviation and a preset positive number is calculated. The load trend regularity is the ratio of the change approximation to the cumulative value; The formula for calculating the load mutation degree is as follows: In the formula, This represents the load mutation degree at the i-th data collection time within the day; , These represent the average load data within the first interval before the i-th collection time on the same day and the previous day, respectively. , represents the dispersion of the load data in the 1st and yth intervals before the i-th collection time within the day, respectively; Y represents the number of intervals at the i-th collection time within the day; It represents the average of the time intervals corresponding to all peak points in the first interval before the i-th collection time on the same day; This indicates the absolute value operation; The process of obtaining the parameter adjustment coefficient is as follows: the load trend regularity is mapped to a first positive number, and the parameter adjustment coefficient is twice the normalized value of the ratio of the load mutation degree to the first positive number; The trend smoothing coefficient at the current moment is the product of the parameter adjustment coefficient and the preset initial trend smoothing coefficient.
2. The microgrid stability control method applied to new energy charging stations as described in claim 1, characterized in that, The method for calculating the approximation of the change is as follows: calculate the product of the accumulated value and the accumulated sum, and the approximation of the change is the ratio of the sum to the product.
3. The microgrid stability control method applied to new energy charging stations as described in claim 1, characterized in that, The method for predicting load data within a preset time period after the current moment is as follows: All load data collected before the current time within the day are arranged in time sequence to form a microgrid load data sequence. The microgrid load data sequence is used as the input of a time series prediction algorithm, and the output is the load data of a preset number of collection times after the current time. The trend smoothing parameter of the current time is used as the smoothing parameter of the trend component in the time series prediction algorithm.
4. The microgrid stability control method applied to new energy charging stations as described in claim 1, characterized in that, The method of controlling the microgrid to connect to or disconnect from the grid based on the prediction results includes: Obtain the output power of the microgrid at the current moment. If the output power is greater than or equal to the load data of all collected moments after the current moment as predicted, switch the microgrid to island mode; otherwise, connect the microgrid to the main grid.
5. A microgrid stability control system applied to new energy charging stations, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the microgrid stability control method for new energy charging stations as described in any one of claims 1-4.