Method and system for coordinated optimization of wind power and photovoltaic energy storage ratio

Abstract

The application discloses a wind power photovoltaic energy storage matching coordination optimization method and system, relates to the wind power photovoltaic joint management and control technical field, carries out feature analysis on historical data, splits and reduces dimension, obtains a characteristic parameter array, and further constructs a mark sub data section set; similar historical characteristic parameter array sequences are combined and compared, highly consistent historical sub data section groups are averaged, an averaged sub data section group is constructed; a comparison sub data section group library is established, used for real-time identification and comparison of wind power generation, photovoltaic power generation and weather information, so that a matched reference sub data section group is determined, and a wind power photovoltaic energy storage matching strategy is regulated and controlled accordingly; the technical scheme disclosed by the application realizes the stability and economy of the wind power photovoltaic energy storage system.

Description

Technical Field

[0001] This invention relates to the field of wind power and photovoltaic joint management and control technology, specifically to a method and system for coordinating and optimizing the ratio of wind power, photovoltaic and energy storage. Background Technology

[0002] With the deepening global utilization of renewable energy, wind and solar power, as clean and renewable energy sources, have seen their grid connection rates rise year by year. However, the output of wind and solar power is strongly influenced by natural conditions (such as wind speed and sunlight), exhibiting significant intermittency and instability, which poses a significant challenge to the stable operation of the power grid. To address this challenge, energy storage technology has emerged as a key means to balance the output fluctuations of wind and solar power and improve the stability and reliability of power supply.

[0003] In wind and solar power systems, energy storage devices not only store excess energy for release when power generation is insufficient, but also undertake multiple tasks such as smoothing power output, peak shaving and valley filling, and improving power quality. However, determining the optimal output ratio between wind and solar power generation and energy storage devices to ensure the stability and economy of power supply has become a pressing technical challenge. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method and system for coordinating and optimizing the ratio of wind power, photovoltaic power, and energy storage, which can effectively solve the aforementioned problems.

[0005] The technical solution adopted in this invention is as follows:

[0006] This invention provides a method for coordinating and optimizing the ratio of wind power, photovoltaic power, and energy storage, comprising:

[0007] Historical wind power generation data, historical photovoltaic power generation data, and historical weather information are acquired, and based on the correspondence of equivalent time, the historical wind power generation data, historical photovoltaic power generation data, and historical weather information are aligned to form a historical data sequence group;

[0008] Feature analysis is performed on the historical data sequence group, and based on the analysis results, the historical data sequence group is split into several historical sub-data segment groups. Feature dimensionality reduction is performed on each historical sub-data segment group to obtain a dimensionality-reduced feature parameter array. The historical sub-data segment groups are combined according to a preset combination method to obtain several flag sub-data segment sets, and several dimensionality-reduced feature parameter arrays corresponding to each flag sub-data segment set are determined, which are denoted as the flag feature parameter array sequence.

[0009] The historical sub-data segments are randomly and dynamically combined according to their original order to obtain several sets of historical sub-data segments. Several dimensionality-reduced feature parameter arrays corresponding to each set of historical sub-data segments are determined and denoted as historical feature parameter array sequences. Several marker feature parameter array sequences are randomly selected, and the difference parameters between the marker feature parameter array sequences and the historical feature parameter array sequences are calculated respectively. The average value of all difference parameters is calculated to obtain the average difference parameter. Based on the average difference parameter, the historical feature parameter array sequences are classified.

[0010] The historical feature parameter array sequences within the same category are randomly combined, and the first degree of consistency between the combined historical feature parameter array sequences is analyzed. If the first degree of consistency is greater than or equal to a preset value, the corresponding historical sub-data segment groups are compared, and the second degree of consistency of the historical sub-data segment groups is calculated. If the second degree of consistency is greater than or equal to a preset value, the historical sub-data segment groups are averaged to obtain averaged sub-data segment groups, and the historical sub-data segment groups corresponding to the averaged sub-data segment groups are removed.

[0011] The remaining historical sub-data segment groups and averaged sub-data segment groups are denoted as the comparison sub-data segment group library. The comparison sub-data segment group library is used to identify and compare real-time wind power generation data, real-time photovoltaic power generation data, and real-time weather information to determine reference sub-data segment groups that match the real-time wind power generation data, real-time photovoltaic power generation data, and real-time weather information. Based on the subsequent data corresponding to the reference sub-data segment groups, the wind power, photovoltaic, and energy storage matching strategy is adjusted.

[0012] In some embodiments disclosed in this invention, aligning historical wind power generation data, historical photovoltaic power generation data, and historical weather information to form a historical data sequence group includes:

[0013] A time axis is constructed, and a parameter axis is constructed in the vertical direction of the time axis. Based on the analysis of historical wind power generation data, historical photovoltaic power generation data, and historical weather information, the historical wind power generation, historical photovoltaic power generation, historical wind intensity, and historical solar intensity at different time points are respectively configured between the time axis and the parameter axis to obtain the historical wind power generation curve, historical photovoltaic power generation curve, historical wind intensity curve, and historical solar intensity curve.

[0014] The historical wind power generation curve, historical photovoltaic power generation curve, historical wind intensity curve, and historical solar irradiance curve are aligned to form a historical data sequence group.

[0015] In some embodiments disclosed in this invention, the step of performing feature analysis on historical data sequence groups and, based on the analysis results, splitting the historical data sequence groups to obtain several historical sub-data segment groups includes:

[0016] Curvature sensitivity thresholds were set for historical wind power generation curves, historical photovoltaic power generation curves, historical wind intensity curves, and historical solar intensity curves, and the points of interest on the curves were determined based on the curvature sensitivity thresholds.

[0017] The analysis involves accumulating features of the points of interest on the curve, including the analysis of the number of points of interest and the analysis of the degree of curvature accumulation. The method for analyzing the number of points of interest includes counting the number of points of interest over time, and resetting the count to zero after each segmentation and truncation of the curve. The method for analyzing the degree of curvature accumulation includes counting the cumulative value of curvature corresponding to each point of interest over time, which is recorded as the degree of curvature accumulation. The degree of curvature accumulation is also reset to zero after each segmentation and truncation of the curve.

[0018] The feature degree of a single curve is calculated in real time based on the number of focus mapping points and the degree of curvature accumulation. Based on the feature degree of all curves, the comprehensive feature degree is determined. If the comprehensive feature degree is greater than or equal to the preset value, all curves are truncated simultaneously, and the combination of all curve segments at this time is recorded as the historical sub-data segment group.

[0019] In some embodiments disclosed in this invention, the expression for calculating the degree of comprehensive features is as follows:

[0020]

[0021] Among them, T zonghe To assess the overall characteristic level, J i is the feature weight adjustment coefficient for the i-th curve. When the curve is a historical wind power generation curve, i is 1; when the curve is a historical photovoltaic power generation curve, i is 2; when the curve is a historical wind intensity curve, i is 3; when the curve is a historical solar intensity curve, i is 4. L1 is the multiplier adjustment coefficient for the number of focus mapping points, g1 is the number of focus mapping points, L2 is the multiplier adjustment coefficient for the curvature accumulation degree, and g2 is the curvature accumulation degree.

[0022] In some embodiments disclosed in this invention, the step of performing feature dimensionality reduction on each historical sub-data segment group to obtain a dimensionality-reduced feature parameter array includes:

[0023] Scan and analyze each curve segment in the historical sub-data segment group to determine the time length d1, the proportion of rising and falling trends d2, and the maximum vertical axis difference d3 corresponding to each curve segment. Then, the time length d1, the proportion of rising and falling trends d2, and the maximum vertical axis difference d3 are identified as the dimensionality-reduced feature parameter group (d1, d2, d3) of the curve segment.

[0024] The combination of the dimensionality-reduced feature parameter groups (d1, d2, d3) corresponding to each curve is denoted as the dimensionality-reduced feature parameter array [d1-i, d2-i, d3-i].

[0025] In some embodiments disclosed in this invention, the calculation of the difference parameters between the flag feature parameter array sequence and the historical feature parameter array sequence includes:

[0026] The marker feature parameter array sequence and the historical feature parameter array sequence are misaligned several times, and the sub-difference parameter is calculated after each misalignment. The method for calculating the sub-difference parameter includes calculating the difference of each feature parameter, and combining the preset weight coefficient of each feature parameter to calculate the single difference parameter corresponding to a single feature parameter array, and calculating the cumulative sum of the single difference parameters corresponding to all feature parameter arrays to obtain the sub-difference parameter of the difference parameter between the marker feature parameter array sequence and the historical feature parameter array sequence.

[0027] The smallest sub-difference parameter is selected and denoted as the difference parameter between the marker feature parameter array sequence and the historical feature parameter array sequence.

[0028] In some embodiments disclosed in this invention, the first degree of consistency among the historical feature parameter array sequences in the analysis combination includes:

[0029] By comparing each relative historical feature parameter array, the sub-match degree is determined. The method for determining the sub-match degree includes calculating the difference of each historical feature parameter and combining it with the preset weight coefficient of each historical feature parameter to calculate the sub-match degree corresponding to a single historical feature parameter array.

[0030] Based on the cumulative sum of sub-matching degrees and the continuous performance characteristics of sub-matching degrees, the first matching degree among the historical feature parameter array sequences is determined;

[0031] The expression for calculating the first degree of similarity is:

[0032]

[0033] Where F represents the first degree of similarity, f xLet δ(n) be the sub-matching degree corresponding to the x-th historical feature parameter array, n be the number of historical feature parameter arrays, δ(n) be the continuity judgment function, used to output the number of historical feature parameter arrays whose sub-matching degree is greater than or equal to the preset value and whose sub-matching degree before and after is greater than or equal to the preset value, K be the continuity influence adjustment coefficient, and b be the continuity influence adjustment constant.

[0034] The expression for calculating the degree of sub-match is:

[0035]

[0036] Where f is the sub-matching degree corresponding to the historical feature parameter array, and h v R represents the difference of the v-th historical feature parameter. v is the weight coefficient of the v-th historical feature parameter, N is the number of historical feature parameters, and H is the preset maximum difference.

[0037] In some embodiments disclosed in this invention, the calculation of the second consistency degree of the historical sub-data segment group includes:

[0038] The curve segments are matched with each other and overlap analysis is performed. The area of ​​difference between the two is calculated, and the second degree of consistency is determined by combining the corresponding consistency influence weight coefficient of the curve segment.

[0039] The expression for calculating the second degree of similarity is:

[0040]

[0041] Where Y represents the second degree of similarity, z represents the preset maximum difference value, and p q S is the consistency influence weighting coefficient for the q-th curve segment. q Let q be the area of ​​difference corresponding to the q-th curve segment, and m be the number of curve segments.

[0042] This invention also provides a system for coordinating and optimizing the ratio of wind power, photovoltaic power, and energy storage, comprising:

[0043] The first module is used to acquire historical wind power generation data, historical photovoltaic power generation data, and historical weather information, and align the historical wind power generation data, historical photovoltaic power generation data, and historical weather information based on the equivalent time correspondence to form a historical data sequence group.

[0044] The second module is used to perform feature analysis on the historical data sequence group, and based on the analysis results, split the historical data sequence group into several historical sub-data segment groups, and perform feature dimensionality reduction on each historical sub-data segment group to obtain a dimensionality-reduced feature parameter array. The historical sub-data segment groups are combined according to a preset combination method to obtain several flag sub-data segment sets, and several dimensionality-reduced feature parameter arrays corresponding to each flag sub-data segment set are determined, which are denoted as the flag feature parameter array sequence.

[0045] The third module is used to randomly and dynamically combine historical sub-data segments according to their original order to obtain several sets of historical sub-data segments, and to determine several dimensionality-reduced feature parameter arrays corresponding to each set of historical sub-data segments, denoted as historical feature parameter array sequences. Several marker feature parameter array sequences are randomly selected, and the difference parameters between the marker feature parameter array sequences and the historical feature parameter array sequences are calculated respectively. The average value of all difference parameters is calculated to obtain the average difference parameter. Based on the average difference parameter, the historical feature parameter array sequences are classified.

[0046] The fourth module is used to randomly combine historical feature parameter array sequences within the same category and analyze the first degree of consistency between the combined historical feature parameter array sequences. If the first degree of consistency is greater than or equal to a preset value, the corresponding historical sub-data segment groups are compared and the second degree of consistency of the historical sub-data segment groups is calculated. If the second degree of consistency is greater than or equal to a preset value, the historical sub-data segment groups are averaged to obtain averaged sub-data segment groups, and the historical sub-data segment groups corresponding to the averaged sub-data segment groups are removed.

[0047] The fifth module is used to record the remaining historical sub-data segment groups and averaged sub-data segment groups as a comparison sub-data segment group library. The comparison sub-data segment group library is used to identify and compare real-time wind power generation data, real-time photovoltaic power generation data and real-time weather information to determine the reference sub-data segment group that matches the real-time wind power generation data, real-time photovoltaic power generation data and real-time weather information. Based on the subsequent data corresponding to the reference sub-data segment group, the wind power, photovoltaic and energy storage matching strategy is adjusted.

[0048] The method and system for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio provided by this invention have the following advantages:

[0049] This invention discloses a method and system for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio, relating to the field of joint wind power and photovoltaic management technology. It involves performing feature analysis on historical data, splitting and reducing its dimensionality to obtain a feature parameter array, and then constructing a set of marker sub-data segments. Similar historical feature parameter array sequences are combined and compared, and highly consistent historical sub-data segment groups are averaged to construct averaged sub-data segment groups. A comparison sub-data segment group library is established for real-time identification and comparison of wind power generation, photovoltaic power generation, and weather information, thereby determining matching reference sub-data segment groups and adjusting the wind power, photovoltaic, and energy storage ratio strategy accordingly. The technical solution disclosed in this invention improves the stability and economy of wind power, photovoltaic, and energy storage systems. Attached Figure Description

[0050] Figure 1 The flowchart shows the method for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio provided by this invention. Detailed Implementation

[0051] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0052] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings and specific embodiments. It should be understood that the preferred embodiments described herein are only for illustration and explanation of the present invention and should not be construed as limiting the scope of protection of the present invention. Those skilled in the art can make some non-essential improvements and adjustments based on the following content of the present invention. In the present invention, unless otherwise expressly specified and limited, the technical terms used in the present invention should have the ordinary meaning understood by those skilled in the art.

[0053] Example:

[0054] This invention discloses a method for coordinating and optimizing the ratio of wind power, photovoltaic power, and energy storage. (See reference...) Figure 1 ,include:

[0055] Step S100: Obtain historical wind power generation data, historical photovoltaic power generation data, and historical weather information, and align the historical wind power generation data, historical photovoltaic power generation data, and historical weather information based on the equivalent time correspondence to form a historical data sequence group.

[0056] In step S100, historical wind power generation data, photovoltaic power generation data, and corresponding historical weather information are first collected. This data forms the basis for subsequent analysis, reflecting the output of wind and photovoltaic power generation under different weather conditions. Based on equivalent time correspondence, these data are aligned, meaning data from the same point in time or the same time period are matched to ensure data synchronization. This resulting historical data sequence provides a complete and ordered dataset for subsequent feature analysis, helping to reveal potential relationships and patterns between the data.

[0057] In some embodiments disclosed in this invention, a method for aligning historical wind power generation data, photovoltaic power generation data, and historical weather information to form a historical data sequence group includes:

[0058] Step S101: Construct a time horizontal axis and a parameter axis for the vertical direction of the time horizontal axis. Based on the analysis of historical wind power generation data, historical photovoltaic power generation data and historical weather information, the historical wind power generation power, historical photovoltaic power generation power, historical wind intensity and historical sunshine intensity at different time nodes are configured between the time horizontal axis and the parameter axis to obtain the historical wind power generation power curve, historical photovoltaic power generation power curve, historical wind intensity curve and historical sunshine intensity curve.

[0059] Step S102: Align the historical wind power generation curve, historical photovoltaic power generation curve, historical wind intensity curve, and historical solar irradiance curve to form a historical data sequence group.

[0060] Step S200: Perform feature analysis on the historical data sequence group, and based on the analysis results, split the historical data sequence group to obtain several historical sub-data segment groups, and perform feature dimensionality reduction on each historical sub-data segment group to obtain a dimensionality-reduced feature parameter array. Combine the historical sub-data segment groups according to a preset combination method to obtain several flag sub-data segment sets, and determine several dimensionality-reduced feature parameter arrays corresponding to each flag sub-data segment set, denoted as the flag feature parameter array sequence.

[0061] Step S200 involves in-depth feature analysis of the historical data sequence group to extract key information representing the data characteristics. Through analysis, outliers, periodic trends, and correlations can be identified. Based on these analysis results, the historical data sequence group is split into several historical sub-data segments. Each historical sub-data segment represents a specific behavioral pattern of the data over a period of time. Next, feature dimensionality reduction is performed on each historical sub-data segment group. This is to reduce the data dimensionality while minimizing information loss, thereby improving data processing efficiency. The resulting historical feature parameter array is more concise but still retains the core information of the data. Finally, these historical sub-data segment groups are combined according to a preset combination method to form several set of labeled sub-data segments, and the corresponding dimensionality-reduced feature parameter array sequence for each set is determined. These sequences will serve as the basis for subsequent data comparison and classification.

[0062] In some embodiments disclosed in this invention, feature analysis is performed on historical data sequence groups, and based on the analysis results, the historical data sequence groups are split to obtain several historical sub-data segment groups, including:

[0063] Step S201: Set curvature sensitive thresholds for the historical wind power generation curve, historical photovoltaic power generation curve, historical wind intensity curve, and historical solar intensity curve respectively, and determine the points of interest on the curves based on the curvature sensitive thresholds.

[0064] First, curvature sensitivity thresholds are set for four key curves: historical wind power generation curve, historical photovoltaic power generation curve, historical wind intensity curve, and historical solar irradiance curve. The curvature sensitivity threshold is a parameter used to identify significant points of change on the curve. When the curvature of the curve exceeds this threshold, it means that a significant change has occurred at that point; this point of change is called the focus point. By setting the curvature sensitivity threshold, these key points of change can be automatically identified, providing a foundation for subsequent feature analysis. Determining the focus points helps capture important features on the curve, such as abrupt changes and inflection points. These features are crucial for understanding the curve's behavior patterns and for subsequent data segmentation.

[0065] Step S202 involves performing an accumulated feature analysis on the focus mapping points on the curve, including an analysis of the number of focus mapping points and an analysis of the degree of curvature accumulation. The method for analyzing the number of focus mapping points includes counting the number of focus mapping points over time, and resetting the count to zero after each segmentation and truncation of the curve. The method for analyzing the degree of curvature accumulation includes counting the cumulative value of curvature corresponding to each focus mapping point over time, which is recorded as the degree of curvature accumulation. The degree of curvature accumulation is reset to zero after each segmentation and truncation of the curve.

[0066] Step S202 performs cumulative feature analysis on the interest mapping points identified in step S201, including interest mapping point quantity analysis and curvature accumulation degree analysis. The interest mapping point quantity analysis is performed by counting the number of interest mapping points appearing on the curve over time. After each curve segmentation and truncation, the count of interest mapping points is reset to zero to allow for independent statistical analysis of the new curve segment. The curvature accumulation degree analysis is performed by calculating the cumulative curvature value corresponding to each interest mapping point. This cumulative value reflects the degree of change in the curve at the interest mapping point. Similarly, after each curve segmentation and truncation, the curvature accumulation degree is also reset to zero. These two analysis methods together constitute a comprehensive feature description of the interest mapping points, providing a basis for subsequent data segmentation.

[0067] Step S203: Calculate the feature degree of a single curve in real time based on the number of focus mapping points and the degree of curvature accumulation. Determine the comprehensive feature degree based on the feature degree of all curves. If the comprehensive feature degree is greater than or equal to a preset value, truncate all curves simultaneously and record the combination of all curve segments at this time as a historical sub-data segment group.

[0068] In step S203, based on the number of attention mapping points and curvature accumulation obtained in step S202, the characteristic degree of a single curve is calculated in real time. Characteristic degree is a quantitative indicator used to evaluate the richness of features of a curve within the current time period. Then, based on the characteristic degrees of all curves, the comprehensive characteristic degree is determined. The comprehensive characteristic degree considers the characteristics of all curves within the current time period and is a comprehensive evaluation indicator. If the comprehensive characteristic degree is greater than or equal to a preset value, it indicates that the data within the current time period has rich features and is an important data segment. Therefore, all curves are simultaneously truncated, and the combination of all curve segments at this time is recorded as a historical sub-data segment group. The resulting historical sub-data segment group contains rich feature information, which is of great significance for subsequent data analysis and the formulation of energy storage allocation strategies. At the same time, by setting preset values ​​to control the frequency and accuracy of data splitting, it can be ensured that the resulting historical sub-data segment group is both representative and not overly lengthy.

[0069] In some embodiments disclosed in this invention, the expression for calculating the degree of comprehensive features is as follows:

[0070]

[0071] Among them, T zonghe To assess the overall characteristic level, J iis the feature weight adjustment coefficient for the i-th curve. When the curve is a historical wind power generation curve, i is 1; when the curve is a historical photovoltaic power generation curve, i is 2; when the curve is a historical wind intensity curve, i is 3; when the curve is a historical solar intensity curve, i is 4. L1 is the multiplier adjustment coefficient for the number of focus mapping points, g1 is the number of focus mapping points, L2 is the multiplier adjustment coefficient for the curvature accumulation degree, and g2 is the curvature accumulation degree.

[0072] In some embodiments disclosed in this invention, the method for performing feature dimensionality reduction on each historical sub-data segment group to obtain a dimensionality-reduced feature parameter array includes:

[0073] Step S204: Scan and analyze each curve segment in the historical sub-data segment group to determine the time length d1, the proportion of rising and falling trends d2, and the maximum vertical axis difference d3 corresponding to each curve segment, and identify the time length d1, the proportion of rising and falling trends d2, and the maximum vertical axis difference d3 as the dimensionality-reduced feature parameter group (d1, d2, d3) of the curve segment.

[0074] Step S205: Combine the dimensionality-reduced feature parameter groups (d1, d2, d3) corresponding to each curve and denote them as the dimensionality-reduced feature parameter array [d1-i, d2-i, d3-i].

[0075] Step S300: The historical sub-data segments are randomly and dynamically combined according to their original order to obtain several sets of historical sub-data segments. Several dimensionality-reduced feature parameter arrays corresponding to each set of historical sub-data segments are determined and denoted as historical feature parameter array sequences. Several flag feature parameter array sequences are randomly selected, and the difference parameters between the flag feature parameter array sequences and the historical feature parameter array sequences are calculated respectively. The average value of all difference parameters is calculated to obtain the average difference parameter. Based on the average difference parameter, the historical feature parameter array sequences are classified.

[0076] In step S300, the historical sub-data segments are first randomly and dynamically combined according to their original order to generate multiple sets of historical sub-data segments. This is done to introduce diversity through randomness, allowing for a more comprehensive consideration of various possibilities in subsequent comparisons and classifications. Next, the dimensionality-reduced feature parameter array sequence corresponding to each historical sub-data segment set is determined; these sequences are called historical feature parameter array sequences. Then, several marker feature parameter array sequences are randomly selected as references, and the difference parameters between them and the historical feature parameter array sequences are calculated. The difference parameters reflect the similarity or difference between the two sets of data. By calculating the average of all difference parameters, the average difference parameter is obtained; this indicator is used to quantify the overall difference between the historical feature parameter array sequences and the marker feature parameter array sequences. Finally, the historical feature parameter array sequences are classified based on the average difference parameter, grouping sequences with high similarity into one category, facilitating subsequent data processing and the formulation of energy storage allocation strategies.

[0077] In some embodiments disclosed in this invention, the method for calculating the difference parameters between the flag feature parameter array sequence and the historical feature parameter array sequence includes:

[0078] Step S301: Perform several misalignment comparisons between the marker feature parameter array sequence and the historical feature parameter array sequence, and calculate the sub-difference parameter after each misalignment comparison. The method for calculating the sub-difference parameter includes calculating the difference of each feature parameter, and combining the preset weight coefficient of each feature parameter to calculate the single difference parameter corresponding to a single feature parameter array, and calculating the cumulative sum of the single difference parameters corresponding to all feature parameter arrays to obtain the sub-difference parameter of the difference parameter between the marker feature parameter array sequence and the historical feature parameter array sequence.

[0079] Step S302: Select the smallest sub-difference parameter, denoted as the difference parameter between the marker feature parameter array sequence and the historical feature parameter array sequence.

[0080] Step S400: Randomly combine the historical feature parameter array sequences within the same category, and analyze the first degree of consistency between the combined historical feature parameter array sequences. If the first degree of consistency is greater than or equal to a preset value, compare the corresponding historical sub-data segment groups and calculate the second degree of consistency of the historical sub-data segment groups. If the second degree of consistency is greater than or equal to a preset value, average the historical sub-data segment groups to obtain averaged sub-data segment groups, and remove the historical sub-data segment groups corresponding to the averaged sub-data segment groups.

[0081] Step S400 further processes the historical feature parameter array sequences to optimize the dataset structure. First, historical feature parameter array sequences within the same category are randomly combined to form multiple combinations. Then, the first degree of similarity between the historical feature parameter array sequences in each combination is analyzed. The first degree of similarity is a quantitative indicator used to evaluate the similarity between sequences. If the first degree of similarity between sequences in a combination is greater than or equal to a preset value, it indicates that these sequences have high similarity. Next, the historical sub-data segment groups corresponding to each of these sequences are compared, and the second degree of similarity between the historical sub-data segment groups is calculated. The second degree of similarity is also a quantitative indicator used to evaluate the similarity between historical sub-data segment groups. If the second degree of similarity is also greater than or equal to a preset value, it indicates that these historical sub-data segment groups have high similarity. In this case, these historical sub-data segment groups are averaged to obtain more representative averaged sub-data segment groups. Finally, the historical sub-data segment groups corresponding to the averaged sub-data segment groups are removed from the dataset to reduce redundant data and improve data processing efficiency.

[0082] In some embodiments disclosed in this invention, the method for analyzing the first degree of consistency between historical feature parameter array sequences in a combination includes:

[0083] Step S401: Compare each relative historical feature parameter array to determine the sub-matching degree. The method for determining the sub-matching degree includes calculating the difference of each historical feature parameter and combining it with the preset weight coefficient of each historical feature parameter to calculate the sub-matching degree corresponding to a single historical feature parameter array.

[0084] Step S402: Based on the cumulative sum of sub-matching degrees and the continuous performance characteristics of sub-matching degrees, determine the first matching degree among the historical feature parameter array sequences.

[0085] The expression for calculating the first degree of similarity is:

[0086]

[0087] Where F represents the first degree of similarity, f x Let δ(n) be the sub-matching degree corresponding to the x-th historical feature parameter array, n be the number of historical feature parameter arrays, δ(n) be the continuity judgment function, used to output the number of historical feature parameter arrays whose sub-matching degree is greater than or equal to the preset value and whose sub-matching degree before and after is greater than or equal to the preset value, K be the continuity influence adjustment coefficient, and b be the continuity influence adjustment constant.

[0088] The expression for calculating the degree of sub-match is:

[0089]

[0090] Where f is the sub-matching degree corresponding to the historical feature parameter array, and h v R represents the difference of the v-th historical feature parameter. v is the weight coefficient of the v-th historical feature parameter, N is the number of historical feature parameters, and H is the preset maximum difference.

[0091] In some embodiments disclosed in this invention, the method for calculating the second consistency degree of a historical sub-data segment group includes:

[0092] Step S403: Match the curve segments to each other and perform overlap analysis. Calculate the area of ​​difference between them and determine the second degree of consistency by combining the corresponding consistency influence weight coefficient of the curve segments.

[0093] The expression for calculating the second degree of similarity is:

[0094]

[0095] Where Y represents the second degree of similarity, z represents the preset maximum difference value, and p q S is the consistency influence weighting coefficient for the q-th curve segment. q Let q be the area of ​​difference corresponding to the q-th curve segment, and m be the number of curve segments.

[0096] Step S500: The set of remaining historical sub-data segment groups and averaged sub-data segment groups is recorded as the comparison sub-data segment group library. The comparison sub-data segment group library is used to identify and compare real-time wind power generation data, real-time photovoltaic power generation data and real-time weather information to determine the reference sub-data segment group that matches the real-time wind power generation data, real-time photovoltaic power generation data and real-time weather information. Based on the subsequent data corresponding to the reference sub-data segment group, the wind power, photovoltaic and energy storage matching strategy is adjusted.

[0097] In step S500, the remaining historical sub-data segment groups and averaged sub-data segment groups are combined to form a comparison sub-data segment group library. This library contains various historical data patterns that can be used for real-time data identification and comparison. When real-time wind power generation data, real-time photovoltaic power generation data, and real-time weather information are acquired, the comparison sub-data segment group library is used for comparison to find the reference sub-data segment group that best matches the real-time data. Since the reference sub-data segment group has a similar historical data pattern to the real-time data, it can be assumed that their data change trends will also be similar in the future. Based on this assumption, the future change trend of the real-time data can be predicted based on the subsequent data corresponding to the reference sub-data segment group, and the wind power, photovoltaic, and energy storage matching strategy can be dynamically adjusted accordingly. This can improve the response speed and adaptability of the energy storage system, ensuring that the energy storage system can operate efficiently and stably.

[0098] In some embodiments disclosed in this invention, a coordinated optimization system for the wind power, photovoltaic, and energy storage ratio is also disclosed, characterized in that it includes:

[0099] The first module is used to acquire historical wind power generation data, historical photovoltaic power generation data, and historical weather information, and align the historical wind power generation data, historical photovoltaic power generation data, and historical weather information based on the equivalent time correspondence to form a historical data sequence group.

[0100] The second module is used to perform feature analysis on the historical data sequence group, and based on the analysis results, split the historical data sequence group into several historical sub-data segment groups, and perform feature dimensionality reduction on each historical sub-data segment group to obtain a dimensionality-reduced feature parameter array. The historical sub-data segment groups are combined according to a preset combination method to obtain several flag sub-data segment sets, and several dimensionality-reduced feature parameter arrays corresponding to each flag sub-data segment set are determined, which are denoted as the flag feature parameter array sequence.

[0101] The third module is used to randomly and dynamically combine historical sub-data segments according to their original order to obtain several sets of historical sub-data segments, and to determine several dimensionality-reduced feature parameter arrays corresponding to each set of historical sub-data segments, denoted as historical feature parameter array sequences. Several marker feature parameter array sequences are randomly selected, and the difference parameters between the marker feature parameter array sequences and the historical feature parameter array sequences are calculated respectively. The average value of all difference parameters is calculated to obtain the average difference parameter. Based on the average difference parameter, the historical feature parameter array sequences are classified.

[0102] The fourth module is used to randomly combine historical feature parameter array sequences within the same category and analyze the first degree of consistency between the combined historical feature parameter array sequences. If the first degree of consistency is greater than or equal to a preset value, the corresponding historical sub-data segment groups are compared and the second degree of consistency of the historical sub-data segment groups is calculated. If the second degree of consistency is greater than or equal to a preset value, the historical sub-data segment groups are averaged to obtain averaged sub-data segment groups, and the historical sub-data segment groups corresponding to the averaged sub-data segment groups are removed.

[0103] The fifth module is used to record the remaining historical sub-data segment groups and averaged sub-data segment groups as a comparison sub-data segment group library. The comparison sub-data segment group library is used to identify and compare real-time wind power generation data, real-time photovoltaic power generation data and real-time weather information to determine the reference sub-data segment group that matches the real-time wind power generation data, real-time photovoltaic power generation data and real-time weather information. Based on the subsequent data corresponding to the reference sub-data segment group, the wind power, photovoltaic and energy storage matching strategy is adjusted.

[0104] This invention discloses a method and system for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio, relating to the field of joint wind power and photovoltaic management technology. It involves performing feature analysis on historical data, splitting and reducing its dimensionality to obtain a feature parameter array, and then constructing a set of marker sub-data segments. Similar historical feature parameter array sequences are combined and compared, and highly consistent historical sub-data segment groups are averaged to construct averaged sub-data segment groups. A comparison sub-data segment group library is established for real-time identification and comparison of wind power generation, photovoltaic power generation, and weather information, thereby determining matching reference sub-data segment groups and adjusting the wind power, photovoltaic, and energy storage ratio strategy accordingly. The technical solution disclosed in this invention improves the stability and economy of wind power, photovoltaic, and energy storage systems.

[0105] Through the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented in hardware or by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) and includes several instructions to cause a computer device (such as a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0106] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for coordinating and optimizing the ratio of wind power, photovoltaic power, and energy storage, characterized in that, include: Historical wind power generation data, historical photovoltaic power generation data, and historical weather information are acquired, and based on the correspondence of equivalent time, the historical wind power generation data, historical photovoltaic power generation data, and historical weather information are aligned to form a historical data sequence group; Feature analysis is performed on the historical data sequence group, and based on the analysis results, the historical data sequence group is split into several historical sub-data segment groups. Feature dimensionality reduction is performed on each historical sub-data segment group to obtain a dimensionality-reduced feature parameter array. The historical sub-data segment groups are combined according to a preset combination method to obtain several flag sub-data segment sets, and several dimensionality-reduced feature parameter arrays corresponding to each flag sub-data segment set are determined, which are denoted as the flag feature parameter array sequence. The historical sub-data segments are randomly and dynamically combined according to their original order to obtain several sets of historical sub-data segments. Several dimensionality-reduced feature parameter arrays corresponding to each set of historical sub-data segments are determined and denoted as historical feature parameter array sequences. Several marker feature parameter array sequences are randomly selected, and the difference parameters between the marker feature parameter array sequences and the historical feature parameter array sequences are calculated respectively. The average value of all difference parameters is calculated to obtain the average difference parameter. Based on the average difference parameter, the historical feature parameter array sequences are classified. The historical feature parameter array sequences within the same category are randomly combined, and the first degree of consistency between the combined historical feature parameter array sequences is analyzed. If the first degree of consistency is greater than or equal to a preset value, the corresponding historical sub-data segment groups are compared, and the second degree of consistency of the historical sub-data segment groups is calculated. If the second degree of consistency is greater than or equal to a preset value, the historical sub-data segment groups are averaged to obtain averaged sub-data segment groups, and the historical sub-data segment groups corresponding to the averaged sub-data segment groups are removed. The remaining historical sub-data segment groups and averaged sub-data segment groups are denoted as the comparison sub-data segment group library. The comparison sub-data segment group library is used to identify and compare real-time wind power generation data, real-time photovoltaic power generation data, and real-time weather information to determine reference sub-data segment groups that match the real-time wind power generation data, real-time photovoltaic power generation data, and real-time weather information. Based on the subsequent data corresponding to the reference sub-data segment groups, the wind power, photovoltaic, and energy storage matching strategy is adjusted.

2. The method for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio according to claim 1, characterized in that, The process of aligning historical wind power generation data, historical photovoltaic power generation data, and historical weather information to form a historical data sequence group includes: A time axis is constructed, and a parameter axis is constructed in the vertical direction of the time axis. Based on the analysis of historical wind power generation data, historical photovoltaic power generation data, and historical weather information, the historical wind power generation, historical photovoltaic power generation, historical wind intensity, and historical solar intensity at different time points are respectively configured between the time axis and the parameter axis to obtain the historical wind power generation curve, historical photovoltaic power generation curve, historical wind intensity curve, and historical solar intensity curve. The historical wind power generation curve, historical photovoltaic power generation curve, historical wind intensity curve, and historical solar irradiance curve are aligned to form a historical data sequence group.

3. The method for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio according to claim 2, characterized in that, The process involves performing feature analysis on historical data sequence groups and, based on the analysis results, splitting the historical data sequence groups to obtain several historical sub-data segment groups, including: Curvature sensitivity thresholds were set for historical wind power generation curves, historical photovoltaic power generation curves, historical wind intensity curves, and historical solar intensity curves, and the points of interest on the curves were determined based on the curvature sensitivity thresholds. The analysis involves accumulating features of the points of interest on the curve, including the analysis of the number of points of interest and the analysis of the degree of curvature accumulation. The method for analyzing the number of points of interest includes counting the number of points of interest over time, and resetting the count to zero after each segmentation and truncation of the curve. The method for analyzing the degree of curvature accumulation includes counting the cumulative value of curvature corresponding to each point of interest over time, which is recorded as the degree of curvature accumulation. The degree of curvature accumulation is also reset to zero after each segmentation and truncation of the curve. The feature degree of a single curve is calculated in real time based on the number of focus mapping points and the degree of curvature accumulation. Based on the feature degree of all curves, the comprehensive feature degree is determined. If the comprehensive feature degree is greater than or equal to the preset value, all curves are truncated simultaneously, and the combination of all curve segments at this time is recorded as the historical sub-data segment group.

4. The method for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio according to claim 3, characterized in that, The expression for calculating the degree of comprehensive features is: ； in, To assess the degree of comprehensive characteristics, For the first The characteristic weight adjustment coefficient of each curve, when the curve is a historical wind power generation curve. The value is 1, when the curve is the historical photovoltaic power generation curve. The value is 2, when the curve is the historical wind intensity curve. The value is 3, when the curve is a historical light intensity curve. It is 4. To focus on the impact of the number of mapping points on the multiplier adjustment factor, To focus on the number of mapping points, The degree of curvature accumulation affects the magnification adjustment factor. This represents the degree of curvature accumulation.

5. The method for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio according to claim 3, characterized in that, The step of performing feature dimensionality reduction on each historical sub-data segment group to obtain a dimensionality-reduced feature parameter array includes: Scan and analyze each curve segment in the historical sub-data segment group to determine the time length d1, the proportion of rising and falling trends d2, and the maximum vertical axis difference d3 corresponding to each curve segment. Then, the time length d1, the proportion of rising and falling trends d2, and the maximum vertical axis difference d3 are identified as the dimensionality-reduced feature parameter group of the curve segment. The combination of the dimensionality-reduced feature parameter groups corresponding to each curve is denoted as the dimensionality-reduced feature parameter array.

6. The method for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio according to claim 5, characterized in that, The difference parameters calculated between the flag feature parameter array sequence and the historical feature parameter array sequence include: The marker feature parameter array sequence and the historical feature parameter array sequence are misaligned several times, and the sub-difference parameter is calculated after each misalignment. The method for calculating the sub-difference parameter includes calculating the difference of each feature parameter, and combining the preset weight coefficient of each feature parameter to calculate the single difference parameter corresponding to a single feature parameter array, and calculating the cumulative sum of the single difference parameters corresponding to all feature parameter arrays to obtain the sub-difference parameter of the difference parameter between the marker feature parameter array sequence and the historical feature parameter array sequence. The smallest sub-difference parameter is selected and denoted as the difference parameter between the marker feature parameter array sequence and the historical feature parameter array sequence.

7. The method for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio according to claim 5, characterized in that, The first degree of consistency among the historical feature parameter array sequences in the analysis combination includes: By comparing each relative historical feature parameter array, the sub-match degree is determined. The method for determining the sub-match degree includes calculating the difference of each historical feature parameter and combining it with the preset weight coefficient of each historical feature parameter to calculate the sub-match degree corresponding to a single historical feature parameter array. Based on the cumulative sum of sub-matching degrees and the continuous performance characteristics of sub-matching degrees, the first matching degree among the historical feature parameter array sequences is determined; The expression for calculating the first degree of similarity is: ； in, For the first degree of consistency, For the first The degree of sub-matching corresponding to each historical feature parameter array The number of historical feature parameter arrays, This is a continuity assessment function used to output the number of historical feature parameter arrays where the sub-match degree is greater than or equal to a preset value, and the sub-match degree before and after the preset value is also greater than or equal to the preset value. For the adjustment factor of continuous influence, Adjustment constants to account for continuity effects; The expression for calculating the degree of sub-match is: ； in, The degree of sub-matching corresponding to the historical feature parameter array. For the first The difference in historical characteristic parameters For the first The weighting coefficients of each historical feature parameter The number of historical characteristic parameters, This is the preset maximum difference.

8. The method for coordinating and optimizing the wind power, photovoltaic, and energy storage ratio according to claim 3, characterized in that, The calculation of the second consistency degree of the historical sub-data segment group includes: The curve segments are matched with each other and overlap analysis is performed. The area of ​​difference between the two is calculated, and the second degree of consistency is determined by combining the corresponding consistency influence weight coefficient of the curve segment. The expression for calculating the second degree of similarity is: ； in, For the second degree of consistency, To preset the maximum difference value, For the first The consistency influence weighting coefficient of each curve segment For the first The area of ​​difference corresponding to each curve segment This represents the number of curve segments.

9. A system for coordinating and optimizing the ratio of wind power, photovoltaic power, and energy storage, characterized in that, include: The first module is used to acquire historical wind power generation data, historical photovoltaic power generation data, and historical weather information, and align the historical wind power generation data, historical photovoltaic power generation data, and historical weather information based on the equivalent time correspondence to form a historical data sequence group. The second module is used to perform feature analysis on the historical data sequence group, and based on the analysis results, split the historical data sequence group into several historical sub-data segment groups, and perform feature dimensionality reduction on each historical sub-data segment group to obtain a dimensionality-reduced feature parameter array. The historical sub-data segment groups are combined according to a preset combination method to obtain several flag sub-data segment sets, and several dimensionality-reduced feature parameter arrays corresponding to each flag sub-data segment set are determined, which are denoted as the flag feature parameter array sequence. The third module is used to randomly and dynamically combine historical sub-data segments according to their original order to obtain several sets of historical sub-data segments, and to determine several dimensionality-reduced feature parameter arrays corresponding to each set of historical sub-data segments, denoted as historical feature parameter array sequences. Several marker feature parameter array sequences are randomly selected, and the difference parameters between the marker feature parameter array sequences and the historical feature parameter array sequences are calculated respectively. The average value of all difference parameters is calculated to obtain the average difference parameter. Based on the average difference parameter, the historical feature parameter array sequences are classified. The fourth module is used to randomly combine historical feature parameter array sequences within the same category and analyze the first degree of consistency between the combined historical feature parameter array sequences. If the first degree of consistency is greater than or equal to a preset value, the corresponding historical sub-data segment groups are compared and the second degree of consistency of the historical sub-data segment groups is calculated. If the second degree of consistency is greater than or equal to a preset value, the historical sub-data segment groups are averaged to obtain averaged sub-data segment groups, and the historical sub-data segment groups corresponding to the averaged sub-data segment groups are removed. The fifth module is used to record the remaining historical sub-data segment groups and averaged sub-data segment groups as a comparison sub-data segment group library. The comparison sub-data segment group library is used to identify and compare real-time wind power generation data, real-time photovoltaic power generation data and real-time weather information to determine the reference sub-data segment group that matches the real-time wind power generation data, real-time photovoltaic power generation data and real-time weather information. Based on the subsequent data corresponding to the reference sub-data segment group, the wind power, photovoltaic and energy storage matching strategy is adjusted.

Images