Energy storage system power curve decomposition method and system

Abstract

The embodiment of the invention provides an energy storage system power curve decomposition method and system, and belongs to the technical field of energy storage. The method comprises the steps that an original power curve is collected, and the working type of a target energy storage system is judged based on the original power curve; matching a corresponding power curve decomposition algorithm based on the corresponding working type; and based on the matched power curve decomposition algorithm, decomposing the original power curve to obtain a square wave decomposition curve corresponding to the original power curve. According to the scheme, the decomposition result of the power curve of the energy storage system has higher interpretation force and application value in analysis and optimization of the energy storage system, and the defects of a traditional method in processing a rapid mutation curve are overcome.

Description

Technical Field

[0001] The present invention relates to the field of energy storage technology, and in particular to a power curve decomposition method for an energy storage system and a power curve decomposition system for an energy storage system. Background Art

[0002] By decomposing the power curve, the sources of power fluctuations in different frequency bands or different time periods can be distinguished. For example, short-term fluctuations may be caused by load changes, while long-term fluctuations may be related to the charging and discharging process of the energy storage device or the state of the external power grid. This helps to better understand the operating characteristics of the energy storage system under different conditions. Power curve decomposition can help analyze the power demand of the energy storage system at different time scales, and then formulate more accurate energy management strategies. For example, it can better arrange the charging and discharging timing of the energy storage system and optimize the economic benefits and operating efficiency of the system. By decomposing the power curve, potential unstable factors can be discovered in advance, such as excessive instantaneous power changes or continuous power fluctuations, so as to take necessary control measures to enhance the stability of the system. Power curve decomposition helps to analyze the charging and discharging mode of the energy storage system, thereby inferring its future performance and life. Long-term and frequent deep charging and discharging will cause damage to the energy storage equipment. By decomposing the curve, the health status of the equipment can be estimated earlier and its service life can be extended. By analyzing the decomposed square wave decomposition curve, abnormal conditions in the operation of the energy storage system can be found. For example, abnormal power fluctuations in a specific frequency band may indicate that a component in the system has failed or needs maintenance.

[0003] It can be seen that the power curve decomposition of energy storage systems may be applied in various scenarios. At present, Fourier transform, as a common signal decomposition method, is widely used in the spectrum analysis of power curves. Fourier transform reveals the frequency components of the power curve by decomposing the original power curve into a linear combination of sine and cosine functions. However, this method has certain limitations when facing non-periodic and non-stationary power curves, especially for sudden changes in charge and discharge states that are common in energy storage systems.

[0004] Specifically, the power curve of the energy storage system often shows sudden changes, such as rapid power fluctuations during charging or discharging. The Fourier transform decomposes the power curve into multiple sine functions and cosine functions. This decomposition method is mainly suitable for stable periodic signals, but its expression ability is limited for non-stationary and sudden power curves. When there are rapid mutations or drastic changes in the power curve, it is difficult for the sine and cosine functions to accurately capture these transient characteristics, resulting in a deviation between the decomposition result and the actual power change.

[0005] In addition, the sine and cosine functions do not completely correspond to the physical characteristics of energy storage devices (such as constant current and constant power charging and discharging), which can easily lead to unclear analysis results in the physical sense and fail to intuitively reflect the actual operating status of the energy storage device. Since the charging and discharging mode of energy storage devices has relatively stable characteristics during operation, and Fourier decomposition ignores these physical characteristics, the decomposition results lack explanatory power in practical applications and cannot provide strong support for system optimization. Therefore, the existing power curve decomposition method based on Fourier transform has certain limitations when dealing with the characteristics of rapid mutations in energy storage power curves, and a more effective decomposition scheme is urgently needed to improve this problem. Summary of the invention

[0006] The purpose of the embodiments of the present invention is to provide a method and system for decomposing a power curve of an energy storage system, so as to at least solve the problem that the power curve decomposition method based on Fourier transform has certain limitations when processing the characteristics of rapid mutation of the energy storage power curve.

[0007] In order to achieve the above-mentioned purpose, the first aspect of the present invention provides a power curve decomposition method for an energy storage system, the method comprising: collecting an original power curve, judging the working type of a target energy storage system based on the original power curve; matching a corresponding power curve decomposition algorithm based on the corresponding working type; and decomposing the original power curve based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve.

[0008] Optionally, the working type of the target energy storage system includes: a continuous charging and discharging type and a static power type during charge and discharge conversion; the matching of the corresponding power curve decomposition algorithm based on the corresponding working type includes: if the working type corresponding to the current original power curve is a continuous charging and discharging type, then the corresponding power curve decomposition algorithm is a Walsh transform decomposition algorithm and / or a sine sign function decomposition algorithm; if the working type corresponding to the current original power curve is a static power type during charge and discharge conversion, then the corresponding power curve decomposition algorithm is a combined decomposition algorithm.

[0009] Optionally, if the currently matched power curve decomposition algorithm is the Walsh transform decomposition algorithm, the original power curve is decomposed based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve, including: determining the amount of sample point data corresponding to the original power curve; performing discretization processing on the original power curve based on the determined amount of sample point data to obtain equally spaced discretized values; constructing an adaptive Hadamard matrix based on the determined amount of sample point data; performing Walsh transform based on the equally spaced discretized values ​​and the constructed Hadamard matrix to obtain the decomposed square wave decomposition curve.

[0010] Optionally, determining the corresponding sample point data volume includes: determining whether there is a sample point data volume set by the user, and if so, using the sample point data volume set by the user as the determined sample point data volume; if not, generating the corresponding sample point data volume based on the original power curve and the time range required for analysis.

[0011] Optionally, the method of generating a corresponding amount of sample point data based on the original power curve and the time range required for analysis includes: determining the time resolution of the original power curve, and dividing it into a predetermined number of sample point data amounts based on the time resolution and the time required for analysis; performing characteristic expressions of the original power curve at each sample point data amount, and taking the predetermined amount of sample point data that can satisfy the description of the power curve change characteristics and has the smallest amount of sample point data as the generated sample point data amount.

[0012] Optionally, the discretization processing is performed on the original power curve based on the determined sample point data volume to obtain equally spaced discretized values, including: determining the domain interval of the original power curve, and determining the starting point and the end point of the domain interval; calculating the interval between each sample point according to the length of the domain interval and the determined number of sample points; starting from the starting point, generating each intermediate sample point one by one based on the calculated interval between each sample point until the last sample point adjacent to the end point is obtained; and obtaining equally spaced discretized values ​​based on the starting point, each intermediate sample point and the end point.

[0013] Optionally, obtaining equally spaced discretized values ​​based on the starting point, each intermediate sample point and the cutoff point includes: mapping the starting point, each intermediate sample point and the cutoff point to the original power curve one by one, and using the function value corresponding to the mapping point as the discretized value at the corresponding point.

[0014] Optionally, constructing an adapted Hadamard matrix based on the determined sample point data volume includes: corresponding the determined sample point data volume as the target order of the Hadamard matrix; starting from the initial matrix, performing recursive processing step by step to obtain higher-order Hadamard matrices in turn until the target order is reached, stopping the recursive processing, and obtaining the adapted Hadamard matrix; wherein the initial matrix is ​​expressed as:

[0015] The recursive processing rules are:

[0016] in, is a Hadamard matrix of order n.

[0017] Optionally, the performing of Walsh transform based on the equally spaced discretized numerical values ​​and the constructed Hadamard matrix to obtain the decomposed square wave decomposition curve includes: performing vectorization processing on the equally spaced discretized numerical values ​​to obtain a discretized numerical vector; performing a Walsh transform rule based on the discretized numerical vector and the constructed Hadamard matrix to obtain a coefficient vector obtained by the Walsh transform; wherein the Walsh transform rule is:

[0018] in, is the coefficient vector obtained by Walsh transform; n is the order of the Hadamard matrix; is a discretized numerical vector; the coefficient vector obtained by the Walsh transform corresponds to the projection of the original power curve on different Walsh basis functions; based on the coefficient vector obtained by the Walsh transform, the decomposed square wave decomposition curve is obtained.

[0019] Optionally, obtaining the decomposed square wave decomposition curve based on the coefficient vector obtained by Walsh transform includes: selecting a corresponding set of Walsh basis functions based on a determined amount of sample point data; multiplying each Walsh basis function with the coefficient vector obtained by Walsh transform to obtain a separate decomposition curve as the decomposed square wave decomposition curve.

[0020] Optionally, if the currently matched power curve decomposition algorithm is a sinusoidal sign function decomposition algorithm, the original power curve is decomposed based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve, including: decomposing the original power curve into multiple sub-curves based on period differences, each sub-curve having a different period; traversing each sub-curve, and assigning a function value of each point on each sub-curve based on a positive and negative relationship; transforming each sub-curve based on the assigned value of each point, and using the obtained transformed curve as a corresponding square wave decomposition curve.

[0021] Optionally, the traversal of each sub-curve and the assignment of the function value of each point on each sub-curve based on a positive and negative relationship include: assigning a point with a positive function value to a first assignment, and assigning a copied point with a second assignment; wherein the absolute value of the first assignment is equal to the absolute value of the second assignment; the first assignment is a positive value, and the second assignment is a negative value.

[0022] Optionally, if the currently matched power curve decomposition algorithm is a combined decomposition algorithm, the original power curve is decomposed based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve, including: decomposing the original power curve into multiple sub-curves based on period differences, each sub-curve having a different period; traversing each sub-curve, assigning a value of 0 to the sub-curve segment in the static power stage, and assigning a function value of each point on the sub-curve segment in the non-static power stage based on a positive and negative relationship.

[0023] Optionally, the function value of each point on the sub-curve segment of the non-stationary power stage is assigned based on a positive and negative relationship, including: assigning a point with a positive function value to a third assignment, and assigning a copied function value to a fourth assignment; wherein the absolute value of the third assignment is equal to the absolute value of the fourth assignment; the third assignment is a positive value, and the fourth assignment is a negative value.

[0024] A second aspect of the present invention provides a power curve decomposition system for an energy storage system, the system comprising: an acquisition unit for acquiring an original power curve and determining a working type of a target energy storage system based on the original power curve; a matching unit for matching a corresponding power curve decomposition algorithm based on a corresponding working type; and a decomposition unit for decomposing the original power curve based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve.

[0025] On the other hand, the present invention provides a computer-readable storage medium having instructions stored thereon, which, when executed on a computer, enables the computer to execute the above-mentioned method for decomposing a power curve of an energy storage system.

[0026] Through the above technical scheme, the scheme of the present invention can more accurately capture the rapidly changing charging and discharging process in the energy storage system, especially the sudden power fluctuations, by decomposing the original power curve into a square wave curve. This decomposition method can better reflect the operating characteristics of the energy storage device under different working conditions. Compared with the traditional Fourier transform, square wave decomposition is more suitable for non-stationary and non-periodic power curves. Due to the characteristics of the square wave, the decomposition result can more intuitively express the actual operating state of the energy storage system, such as the constant current and constant power charging and discharging process, so that it is closer to the actual situation of the energy storage system in a physical sense. In addition, square wave decomposition can help identify abnormal power fluctuations in the energy storage system, provide more effective control and optimization strategies, and enhance the stability and economic benefits of the system. Therefore, the power curve decomposition technology based on square waves can overcome the limitations of traditional methods in dealing with rapid mutation characteristics, and provide more accurate support for energy management and health monitoring of energy storage systems.

[0027] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings: Figure 1 It is a flowchart of the steps of a method for decomposing a power curve of an energy storage system provided by an embodiment of the present invention; Figure 2 It is a decomposition schematic diagram of performing power curve decomposition based on a sine sign function decomposition algorithm provided by an embodiment of the present invention; Figure 3 It is a decomposition schematic diagram of performing power curve decomposition based on a combined decomposition algorithm provided by an embodiment of the present invention; Figure 4 It is a system structure diagram of a power curve decomposition system for an energy storage system provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0029] The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention.

[0030] Figure 1 FIG. 1 is a flow chart of a method for decomposing a power curve of an energy storage system provided by an embodiment of the present invention. Figure 1 As shown, an embodiment of the present invention provides a method for decomposing a power curve of an energy storage system, the method comprising: Step S10: Collecting the original power curve, and determining the working type of the target energy storage system based on the original power curve.

[0031] Specifically, in the operation of the energy storage system, different working types show different power curve characteristics. In order to improve the operating efficiency of the energy storage system, collecting and analyzing the original power curve is a crucial step. By analyzing these raw data, the working type of the target energy storage system can be determined, and the most suitable power curve decomposition algorithm can be selected accordingly. The working types of energy storage systems are generally divided into two types: continuous charging and discharging type and static power type during charging and discharging conversion.

[0032] Furthermore, for the continuous charging and discharging type, the power curve of the system shows that there is no long-term state of zero power, which means that the system is almost always performing charging and discharging operations. This type of power curve is often manifested in continuous energy input or output, which is suitable for uninterrupted load demand or continuous grid balancing applications. In this case, the power curve decomposition algorithm needs to be able to capture the characteristics of continuous fluctuations, especially to be able to identify small power changes to ensure the smoothness of charging and discharging operations. The use of a suitable decomposition algorithm can help optimize the timing of charging or discharging, avoid system overload or over-discharge, and thus improve the economic benefits and operating life of the energy storage system.

[0033] Furthermore, there is a static power type during the charge and discharge conversion, that is, the system has a long period of static state with zero power between the charging and discharging operations. This usually occurs when the energy storage system is performing battery repair, maintenance or energy balance, and the system is in standby or low-load operation. In this case, the power curve may show sudden changes, such as suddenly entering a large-scale charging or discharging operation from a static state. When decomposing these types of power curves, the algorithm needs to have a high sensitivity to accurately capture these mutation points and provide more accurate charging and discharging strategies.

[0034] By selecting decomposition algorithms for different types of work, not only can the energy management of the energy storage system be optimized, but potential abnormal situations can also be identified in advance. For example, in the continuous charge and discharge type, the algorithm can detect excessively frequent deep charge and discharge, which may cause loss of energy storage equipment and affect its long-term performance. For the situation where there is static power during charge and discharge conversion, the algorithm can help identify sudden abnormal power changes after long-term stagnation, and warn of potential failures or equipment health problems.

[0035] Step S20: matching a corresponding power curve decomposition algorithm based on the corresponding working type.

[0036] Specifically, if the working type corresponding to the current original power curve is a continuous charge and discharge type, the corresponding power curve decomposition algorithm is a Walsh transform decomposition algorithm and / or a sine sign function decomposition algorithm; if the working type corresponding to the current original power curve is a static power type during charge and discharge conversion, the corresponding power curve decomposition algorithm is a combined decomposition algorithm.

[0037] Step S30: based on the matched power curve decomposition algorithm, the original power curve is decomposed to obtain a square wave decomposition curve corresponding to the original power curve.

[0038] Embodiment 1: If the currently matched power curve decomposition algorithm is the Walsh transform decomposition algorithm, the original power curve is decomposed based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve, including: determining the amount of sample point data corresponding to the original power curve; performing discretization processing on the original power curve based on the determined amount of sample point data to obtain equally spaced discretized values; constructing an adaptive Hadamard matrix based on the determined amount of sample point data; performing Walsh transform based on the equally spaced discretized values ​​and the constructed Hadamard matrix to obtain the decomposed square wave decomposition curve.

[0039] Specifically, determining the corresponding sample point data volume includes: judging whether there is a sample point data volume set by the user, and if so, using the sample point data volume set by the user as the determined sample point data volume; if not, generating the corresponding sample point data volume based on the original power curve and the time range required for analysis.

[0040] Furthermore, the method of generating a corresponding amount of sample point data based on the original power curve and the time range required for analysis includes: determining a time resolution of the original power curve, and dividing it into a predetermined number of sample point data amounts based on the time resolution and the time required for analysis; performing characteristic expressions of the original power curve at each sample point data amount, and using a predetermined amount of sample point data that can satisfy the description of the power curve change characteristics and has the smallest amount of sample point data as the generated sample point data amount.

[0041] In the embodiment of the present invention, in the process of decomposition and analysis of the power curve, the selection of the amount of sample point data is crucial and directly affects the accuracy and computational efficiency of the decomposition result. By collecting the original power curve and determining the amount of adapted sample point data, a good foundation can be laid for subsequent discretization processing and Walsh transformation. In the specific implementation, the determination of the amount of sample point data is divided into two cases: if the user sets the amount of sample point data, the value set by the user will be directly adopted; if the amount of data is not set by the user, it is necessary to automatically generate an appropriate amount of sample point data according to the characteristics of the original power curve and the time range required for analysis.

[0042] Furthermore, when there is no user-set sample point data volume, the time resolution of the original power curve must first be analyzed. Time resolution refers to the number of data points collected per unit time in the power curve, which is usually determined by the sampling frequency. High time resolution means a higher sampling frequency, which can capture more subtle changes in the power curve. Next, multiple preset sample point data volumes are generated based on the time resolution and the time range to be analyzed. This process ensures that different numbers of sample points can meet different time ranges and curve change characteristics. Under each sample point data volume, the original power curve is characterized and its ability to describe the curve change characteristics is evaluated. In order to ensure the efficiency and accuracy of the decomposition, a preset value that can accurately describe the power curve change characteristics and has the smallest sample point data volume is selected. This step is very important because too much sample point data volume will increase the computational complexity and data storage overhead, while too little sample point data volume may not accurately capture the sudden changes and dynamic characteristics of the curve. Therefore, when evaluating the descriptive ability of the curve characteristics, a balance point needs to be found, that is, the minimum sample point data volume is used to accurately describe the power curve characteristics.

[0043] Based on the scheme of the present invention, when the user has specific needs, the amount of sample point data can be directly set, which provides the user with operational freedom and personalized selection space. When the user does not provide the amount of data, the optimal amount of sample point data can be automatically generated according to the time resolution and time range of the curve, realizing automatic optimization selection. This automatic selection mechanism greatly improves the efficiency and accuracy of data analysis. In the actual energy storage power curve, the power change is often nonlinear and non-periodic, especially the mutation phenomenon that exists during the charging and discharging process, which is difficult to fully express with a fixed number of sample points. The scheme can ensure the accurate capture and description of the curve by dynamically generating an adaptive amount of sample point data, and can effectively deal with both subtle dynamic changes and large power mutations.

[0044] Furthermore, the automatic optimization of the sample point data volume can not only reduce the storage requirements for redundant data, but also significantly speed up the processing speed of the subsequent Walsh transform. Especially in large-scale energy storage, reducing the computational complexity is of great significance for real-time performance and energy management. Accurately selected sample point data volume helps to improve the physical meaning of the decomposed power curve. The traditional Fourier transform is prone to losing information when faced with rapid mutations because it cannot well express non-stationary power curves. Through the Walsh transform combined with the optimal sample point data volume, the decomposed curve can not only accurately express the changes in the power curve, but also be closer to the physical characteristics of the energy storage device (such as constant current and constant power charging and discharging states), making the analysis results more explanatory.

[0045] In one possible implementation, when performing the decomposition of the power curve, the selection of the number of sample points is critical to the accuracy and efficiency of the calculation. In order to ensure compatibility with the Walsh Transform and the Hadamard Matrix, the optimal number of sample points is usually selected as a power of 2 (e.g., 8, 16, 32, etc.). This selection enables the Hadamard matrix to efficiently represent the decomposition basis functions in the form of orthogonal square waves and ensures the simplicity and speed of the calculation process. Because the size of the Hadamard matrix is ​​a 2n×2n square matrix, if the number of sample points is a power of 2, the Walsh transform can be directly applied to achieve fast matrix operations, greatly reducing the complexity of the calculation.

[0046] Specifically, the domain interval of the original power curve is determined, and the starting point and the end point of the domain interval are determined; the interval between each sample point is calculated according to the length of the domain interval and the determined number of sample points; starting from the starting point, each intermediate sample point is generated one by one based on the calculated interval between each sample point until the last sample point adjacent to the end point is obtained; and equally spaced discretized values ​​are obtained based on the starting point, each intermediate sample point and the end point.

[0047] Furthermore, obtaining equally spaced discretized values ​​based on the starting point, each intermediate sample point and the cutoff point includes: mapping the starting point, each intermediate sample point and the cutoff point to the original power curve one by one, and using the function value corresponding to the mapping point as the discretized value at the corresponding point.

[0048] In the embodiment of the present invention, discretization is a key step in the analysis and processing of the power curve. The power curve is usually presented in a continuous form, and directly using such a curve for analysis may lead to low computational efficiency, especially when dealing with complex mathematical transformations. Therefore, discretization processing is performed on the original power curve based on a determined amount of sample point data, and the continuous curve is converted into a discrete set of numerical points, which is a prerequisite step for power curve decomposition.

[0049] Furthermore, the domain interval of the original power curve is determined, that is, the horizontal axis range of the curve, which is usually expressed as a time range. In this process, it is necessary to clarify the starting point and the end point of the curve, which will serve as the time axis boundary conditions of the power curve. Then, according to the length of the domain interval and the predetermined number of sample points, the interval between each sample point is calculated.

[0050] After calculating the interval between each sample point, the discretization process starts from the starting point of the domain and generates each intermediate sample point one by one until it reaches the last sample point adjacent to the cutoff point. These sample points are equally spaced and cover the entire domain interval, thus ensuring that each part of the original curve is sampled accordingly.

[0051] In order to obtain complete discretized values, it is also necessary to map these sample points back to the original power curve. Specifically, the generated starting point, intermediate sample points, and cutoff points correspond to the corresponding function values ​​on the original power curve, that is, the power value at each discrete sample point. In this way, through one-to-one mapping, the original power curve is converted into a set of equally spaced discrete numerical sequences, which are the basis for subsequent processing (such as Walsh transform or other mathematical transformations).

[0052] Based on the scheme of the present invention, through such discretization processing, the continuous original power curve can be effectively converted into a set of discrete points, which is convenient for further mathematical operations and decomposition. This discretization process not only improves the efficiency of calculation, but also ensures that the characteristics of the power curve are preserved. The generation of equally spaced sample points ensures the uniformity of the data, making the subsequent decomposition results more accurate.

[0053] At the same time, discretization can significantly reduce the amount of data, thereby reducing computational complexity, especially in large-scale data analysis, which can save storage space and speed up calculations. In addition, by mapping the function values ​​of the original curve, these discrete points can truly reflect the power changes at each point of the curve, retaining the important characteristics of the curve and ensuring that the decomposed results have physical meaning. Especially when dealing with the charging and discharging process of the energy storage system, discretization can accurately capture the mutations and details of the power curve, ensuring that the subsequent analysis results are of practical value for the optimization of the energy storage system.

[0054] Specifically, the determined sample point data amount is used as the target order of the Hadamard matrix; starting from the initial matrix, recursive processing is performed step by step to obtain higher-order Hadamard matrices in turn until the target order is reached, and the recursive processing is stopped to obtain an adapted Hadamard matrix; wherein the initial matrix is ​​expressed as:

[0055] The recursive processing rules are:

[0056] in, is a Hadamard matrix of order n.

[0057] In the embodiment of the present invention, it is first necessary to construct a Hadamard matrix of the corresponding order according to the determined amount of sample point data. The order of the Hadamard matrix is ​​usually a power of 2. For example, when the number of sample points is 8, the required Hadamard matrix is ​​an 8×8 matrix. The target order corresponds to the amount of sample point data, which is an important condition to ensure that the Hadamard matrix can effectively perform signal decomposition.

[0058] Based on the scheme of the present invention, by constructing a Hadamard matrix that matches the number of samples, it can be ensured that the subsequent Walsh transform perfectly matches the basis function of the discretized power curve, thereby achieving efficient matrix operations. The recursive construction method of the Hadamard matrix can not only simplify the calculation, but also ensure the orthogonal characteristics of the matrix, which is very important for accurately decomposing the various components in the power curve. Especially when the number of sample points is large, the recursive construction of the Hadamard matrix can greatly reduce the complexity of the calculation and improve the processing speed. In addition, the Hadamard matrix matches the number of sample points well, and can accurately reflect the characteristics in the power curve, whether it is a sudden change or a subtle change, which can be accurately captured by the decomposed matrix basis function. This efficient and accurate decomposition method has extremely high practical value for power analysis of energy storage systems, and can provide strong support for system optimization and operation.

[0059] Specifically, vectorization processing is performed on the equally spaced discretized numerical values ​​to obtain a discretized numerical vector; the Walsh transformation rule is performed based on the discretized numerical vector and the constructed Hadamard matrix to obtain a coefficient vector obtained by the Walsh transformation; wherein the Walsh transformation rule is:

[0060] in, is the coefficient vector obtained by Walsh transform; n is the order of the Hadamard matrix; is a discretized numerical vector; the coefficient vector obtained by the Walsh transform corresponds to the projection of the original power curve on different Walsh basis functions; based on the coefficient vector obtained by the Walsh transform, the decomposed square wave decomposition curve is obtained.

[0061] Furthermore, the decomposition curve of the square wave is obtained based on the coefficient vector obtained by Walsh transform, including: selecting a corresponding set of Walsh basis functions based on a determined amount of sample point data; and multiplying each Walsh basis function with the coefficient vector obtained by Walsh transform to obtain a separate decomposition curve as the decomposition curve of the square wave.

[0062] In an embodiment of the present invention, in the analysis process of the power curve, Walsh transform is an effective signal decomposition method. By performing Walsh transform based on equally spaced discretized values ​​and constructed Hadamard matrix, the complex original power curve can be decomposed into a linear combination of multiple basis functions. Specifically, it is first necessary to vectorize the discretized values, so that a discretized numerical vector can be formed for operation with the Hadamard matrix. The generation of the discretized numerical vector is based on the previously determined sample point data amount, and each discrete point of the curve has a corresponding power value through equally spaced sampling.

[0063] Furthermore, based on the discretized numerical vector and the constructed Hadamard matrix, the matrix-vector multiplication operation is performed according to the rules of Walsh transform. The essence of Walsh transform is to decompose the power curve into different Walsh basis functions, which are orthogonal functions corresponding to the rows of the Hadamard matrix. Through matrix operations, the coefficient vectors after Walsh transform are obtained. These coefficient vectors are the projections of the original power curve on each Walsh basis function, reflecting the decomposition characteristics of the original curve at different frequencies and time scales.

[0064] Furthermore, after obtaining the coefficient vector, the next step is to reconstruct the decomposed square wave decomposition curve based on these coefficients. To this end, it is necessary to select the corresponding set of Walsh basis functions based on the determined amount of sample point data. Each basis function represents the characteristics of the power curve in a certain dimension. Then, each Walsh basis function is multiplied by the corresponding Walsh transform coefficient to obtain a separate decomposition curve. Each decomposition curve represents a component of the original power curve.

[0065] Based on the solution of the present invention, the original power curve can be decomposed into multiple sub-curves with different characteristics through Walsh transform. Each Walsh basis function represents the power change of different frequency or time characteristics, and the coefficient vector quantifies the projection of the power curve on each basis function. Compared with the traditional Fourier transform, the Walsh transform uses a square wave basis function, which can more accurately handle non-periodic changes in the power curve, especially non-stationary signals such as rapid charging and discharging mutations commonly seen in energy storage systems. The Walsh transform provides a more flexible signal decomposition method that can capture subtle changes and mutation points in the power curve. This is particularly important for the optimization of energy storage systems, and can help analyze the switching and transient characteristics of the charging and discharging states. Secondly, based on the coefficient vector obtained by discretization processing and Walsh transform, the components of the original curve can be more clearly analyzed through multiple decomposition curves. Different decomposition curves can reveal the characteristics of different frequency bands in the power curve, thereby achieving a deeper understanding of the curve.

[0066] Embodiment 2: If the currently matched power curve decomposition algorithm is a sine sign function decomposition algorithm, the original power curve is decomposed based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve, including: decomposing the original power curve into multiple sub-curves based on period differences, each sub-curve having a different period; traversing each sub-curve, and assigning a function value of each point on each sub-curve based on a positive and negative relationship; transforming each sub-curve based on the assignment of each point, and using the obtained transformed curve as a corresponding square wave decomposition curve.

[0067] Furthermore, the traversal of each sub-curve and the assignment of the function value of each point on each sub-curve based on the positive and negative relationship include: assigning a point with a positive function value to a first assignment, and assigning a copied point with a second assignment; wherein the absolute value of the first assignment is equal to the absolute value of the second assignment; the first assignment is a positive value, and the second assignment is a negative value.

[0068] In the implementation of the present invention, in the power curve analysis, the sine sign function decomposition algorithm provides an effective means for decomposing the original power curve, especially when dealing with complex, multi-periodic power fluctuations, useful feature information can be extracted from it. The algorithm decomposes the periodic changes of the original power curve and symbolizes the decomposed sub-curves to finally generate the corresponding square wave decomposition curve.

[0069] Furthermore, for the decomposition process of the original power curve, the algorithm will decompose the original curve into multiple sub-curves according to different periodic characteristics. This means that each sub-curve represents the change pattern of the original power curve within a certain period range, so that various fluctuation characteristics in the power curve can be captured finely. Fluctuations in different periods may reflect the dynamics of the energy storage system on different time scales. For example, short-term fluctuations may be related to load changes, while long-term fluctuations may originate from the charging and discharging process of the energy storage system or fluctuations in the external power grid. This decomposition method helps system operation and maintenance personnel better understand the operating conditions of the energy storage system within different time ranges.

[0070] Next, the algorithm traverses each sub-curve and assigns symbolic values ​​to the function values ​​of each point on each sub-curve according to its positive or negative value. Specifically, when the function value of a point is positive, it is assigned the first value (for example, 1), and when the function value is negative, it is assigned the second value (for example, -1). This assignment method ensures that the sign difference of each point is accurately retained, while the first assignment and the second assignment remain equal in absolute value, but represent positive and negative relationships respectively. This method enables the decomposed square wave curve to clearly distinguish between positive and negative power changes, and intuitively reflect the different states of the energy storage system during the charging and discharging process. For example, when the system is in charging mode, the curve may appear as a positive value; while in discharge mode, the curve appears as a negative value.

[0071] Furthermore, by assigning values ​​to each sub-curve and transforming it, the resulting square wave decomposition curve can capture the instantaneous changes and mutations in the power curve of the energy storage system. The advantage of this method is that the square wave decomposition curve simplifies the complex fluctuations in the original power curve, allowing it to be presented in a clearer form, which facilitates the subsequent energy management and control strategy formulation. Especially when dealing with sudden charging and discharging states, the square wave decomposition curve is significantly better than the traditional continuous function expression because the instantaneous jump characteristics of the square wave can more accurately reflect the operating state of the energy storage system during rapid charging and discharging.

[0072] In a possible implementation, Figure 2 As shown in the figure, the bounded period of the sine function is used to decompose the sign function series to obtain sign functions of different periods (frequencies). The processing rules are as follows: .

[0073] Embodiment three: If the currently matched power curve decomposition algorithm is a combined decomposition algorithm, the original power curve is decomposed based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve, including: decomposing the original power curve into multiple sub-curves based on period differences, and each sub-curve has a different period; traversing each sub-curve, assigning a value of 0 to the sub-curve segment in the static power stage, and assigning a function value of each point on the sub-curve segment in the non-static power stage based on a positive and negative relationship.

[0074] Furthermore, the function values ​​of each point on the sub-curve segment of the non-stationary power stage are assigned based on the positive and negative relationship, including: assigning points with positive function values ​​to the third assignment, and assigning points with copied function values ​​to the fourth assignment; wherein the absolute value of the third assignment is equal to the absolute value of the fourth assignment; the third assignment is a positive value, and the fourth assignment is a negative value.

[0075] In the embodiment of the present invention, in the power curve analysis, the combined decomposition algorithm provides a more flexible and accurate way to process complex power fluctuation data, especially in the charge-discharge conversion and static phases that often occur in the operation of the energy storage system. The algorithm decomposes the periodicity of the original power curve, and combines the different characteristics of the static phase and the non-static phase to perform specific assignment processing on each sub-curve segment, thereby generating an accurate square wave decomposition curve.

[0076] Furthermore, the combined decomposition algorithm decomposes the original curve into multiple sub-curves according to the periodic differences of the power curve. These sub-curves correspond to different periodic segments, which can help analyze power fluctuations on different time scales. For example, a short-period sub-curve may correspond to a short-term power fluctuation of the energy storage system, which is often caused by instantaneous load changes or grid instability; while a long-period sub-curve may reflect the long-term behavior of the system, such as the charging and discharging cycle of the energy storage device. This decomposition method helps to deeply understand the operating characteristics of the system at various time scales. Next, the algorithm traverses each sub-curve and processes the curve segments differently according to the static phase and the non-static phase. In the static power phase, the energy storage system is in a relatively static state, and the power curve remains basically unchanged. At this time, the algorithm assigns all points in the static phase to 0, indicating that there is no significant power fluctuation in this phase. In this way, the power fluctuations during charging and discharging can be highlighted in the square wave decomposition curve, and unnecessary static power information can be filtered out.

[0077] Furthermore, in the non-stationary power stage, that is, when the energy storage system is charging or discharging, the algorithm assigns the function value of each point based on the positive and negative relationship. Specifically, the point with a positive function value is assigned the third assignment (for example, 1), and the point with a negative function value is assigned the fourth assignment (for example, -1), and the absolute values ​​of the third assignment and the fourth assignment are the same. This processing method ensures that the power fluctuations in the non-stationary stage are accurately captured, and the square wave decomposition curve can clearly reflect the charging and discharging process of the energy storage system.

[0078] The advantage of the combined decomposition algorithm is that it can take into account the different power characteristics of the static and non-static phases at the same time, thereby generating a more accurate square wave decomposition curve. Through this decomposition method, the power curve of the energy storage system is simplified and optimized under complex and changing conditions. For example, in the operation of the energy storage system, the static power phase often does not affect the overall energy management. Therefore, by assigning the power value of this phase to 0, the operation and maintenance personnel can focus on the power fluctuations in the non-static phase and improve the efficiency and accuracy of the analysis. In the non-static phase, the square wave decomposition curve can not only reflect the direction of power change, but also help identify potential abnormal fluctuations, such as abnormal power mutations or long-term overcharging / discharging operations.

[0079] like Figure 3 , showing a periodic pulse curve with obvious active and static intervals. On the left side of the figure, we see a series of high-frequency pulses that change rapidly between +1 and -1, forming a typical charging and discharging pattern. These pulses represent the exchange of energy during the cycle time. Figure 3 There is a section of the middle curve that is completely flat, with the vertical axis value at zero, indicating that there is no energy conversion or exchange during this period. This static state represents a waiting or buffering stage for the system, and is a transition between two active states of the system. Subsequently, the curve enters an active charge and discharge cycle again, similar to the previous cycle, with pulses varying between +1 and -1. This repetitive pattern indicates that the system has resumed the energy exchange process after a static rest or reset phase.

[0080] In one possible application, the original power curve is decomposed into multiple decomposition curves based on square waves. These decomposition curves represent the characteristics of the power curve in different frequency bands or specific time ranges. Therefore, each decomposition curve can be regarded as the power demand of the system at a specific time or frequency. This feature allows the energy storage system to configure the corresponding energy storage device according to the power demand of each decomposition curve and provide the corresponding power output.

[0081] Furthermore, the pulse form of each decomposition curve also means that they can correspond well to the charge and discharge cycle of the energy storage system. By analyzing the characteristics of these decomposition curves (such as frequency, amplitude and duration), the configuration parameters of the energy storage system, including capacity, power and duration, can be accurately determined. For example, a higher frequency decomposition curve may correspond to a short-term, high-power demand, which requires the energy storage device to be able to provide a short-term high-power output. The low-frequency decomposition curve corresponds to a stable power demand over a longer period of time, which requires a larger capacity but lower power energy storage device.

[0082] Figure 4 FIG. 1 is a structural diagram of a power curve decomposition system for an energy storage system provided by an embodiment of the present invention. Figure 4 As shown, an embodiment of the present invention provides a power curve decomposition system for an energy storage system, the system comprising: a collection unit, used to collect an original power curve, and judge the working type of a target energy storage system based on the original power curve; a matching unit, used to match a corresponding power curve decomposition algorithm based on the corresponding working type; a decomposition unit, used to decompose the original power curve based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve.

[0083] The embodiment of the present invention further provides a computer-readable storage medium, on which instructions are stored, which, when executed on a computer, enable the computer to execute the above-mentioned method for decomposing the power curve of an energy storage system.

[0084] Those skilled in the art will understand that all or part of the steps in the method for implementing the above-mentioned embodiments can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including several instructions for making a single-chip microcomputer, a chip or a processor execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0085] The optional embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above embodiments. Within the technical concept of the embodiments of the present invention, the technical scheme of the embodiments of the present invention can be subjected to a variety of simple modifications, and these simple modifications all belong to the protection scope of the embodiments of the present invention. It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the embodiments of the present invention will not further describe various possible combinations.

[0086] In addition, various embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the embodiments of the present invention, they should also be regarded as the contents disclosed in the embodiments of the present invention.

Claims

1. A method for decomposing a power curve of an energy storage system, characterized in that: The method comprises: Collect the original power curve and determine the working type of the target energy storage system based on the original power curve; Based on the corresponding work type, match the corresponding power curve decomposition algorithm; Based on the matched power curve decomposition algorithm, the original power curve is decomposed to obtain a square wave decomposition curve corresponding to the original power curve.

2. The method according to claim 1, characterized in that The target energy storage system operation types include: There are continuous charge and discharge types and static power types when switching between charge and discharge; The matching of the corresponding power curve decomposition algorithm based on the corresponding work type includes: If the working type corresponding to the current original power curve is a continuous charge and discharge type, the corresponding power curve decomposition algorithm is a Walsh transform decomposition algorithm and / or a sine sign function decomposition algorithm; If the working type corresponding to the current original power curve is a static power type during charge-discharge conversion, the corresponding power curve decomposition algorithm is a combined decomposition algorithm.

3. The method according to claim 2, characterized in that If the currently matched power curve decomposition algorithm is the Walsh transform decomposition algorithm, the original power curve is decomposed based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve, including: Determine the amount of sample point data corresponding to the original power curve; Based on the determined amount of sample point data, the original power curve is discretized to obtain equally spaced discretized values; Constructing an adapted Hadamard matrix based on the determined amount of sample point data; Walsh transform is performed based on the equally spaced discretized values ​​and the constructed Hadamard matrix to obtain the decomposed square wave decomposition curve.

4. The method according to claim 3, characterized in that: Determine the corresponding sample point data volume, including: Determine whether there is a sample point data volume set by the user, and if so, use the sample point data volume set by the user as the determined sample point data volume; If it does not exist, the corresponding sample point data volume is generated based on the original power curve and the time range required for analysis.

5. The method according to claim 4, characterized in that The generating of the corresponding sample point data volume based on the original power curve and the time range to be analyzed includes: Determine the time resolution of the original power curve, and divide it into a predetermined number of sample point data amounts based on the time resolution and the time required for analysis; The original power curve feature expression is performed at each sample point data amount, and the predetermined sample point data amount that can meet the power curve change feature description and has the smallest sample point data amount is used as the generated sample point data amount.

6. The method according to claim 3, characterized in that The discretization processing is performed on the original power curve based on the determined sample point data amount to obtain equally spaced discretized values, including: Determine the definition domain interval of the original power curve, and determine the starting point and the end point of the definition domain interval; According to the length of the domain interval and the determined number of sample points, the interval between each sample point is calculated; Starting from the starting point, each intermediate sample point is generated one by one based on the interval between each sample point obtained by calculation, until the last sample point adjacent to the cutoff point is obtained; Obtain equally spaced discretized values ​​based on the starting point, each intermediate sample point, and the cutoff point.

7. The method according to claim 6, characterized in that The step of obtaining equally spaced discretized values ​​based on the starting point, each intermediate sample point and the cutoff point includes: The starting point, each intermediate sample point and the cutoff point are mapped to the original power curve one by one, and the function value corresponding to the mapping point is used as the discretized value at the corresponding point.

8. The method according to claim 3, characterized in that Based on the determined amount of sample point data, an adapted Hadamard matrix is ​​constructed, including: The determined amount of sample point data is used as the target order of the Hadamard matrix; Starting from the initial matrix, recursive processing is performed step by step to obtain higher-order Hadamard matrices in turn until the target order is reached, and the recursive processing is stopped to obtain the adapted Hadamard matrix; wherein, The initial matrix is ​​expressed as: The recursive processing rules are: in, is a Hadamard matrix of order n.

9. The method according to claim 8, characterized in that The Walsh transform is performed based on the equally spaced discretized values ​​and the constructed Hadamard matrix to obtain a decomposed square wave decomposition curve, including: Performing vectorization processing on the equally spaced discretized numerical values ​​to obtain a discretized numerical vector; Walsh transform rule is performed based on the discretized numerical vector and the constructed Hadamard matrix to obtain the coefficient vector obtained by Walsh transform; where, The Walsh transformation rule is: in, is the coefficient vector obtained by Walsh transformation; n is the order of the Hadamard matrix; is a discretized numerical vector; The coefficient vector obtained by the Walsh transform corresponds to the projection of the original power curve on different Walsh basis functions; Based on the coefficient vector obtained by Walsh transform, the decomposed square wave decomposition curve is obtained.

10. The method according to claim 9, characterized in that The coefficient vector obtained based on the Walsh transform is used to obtain the decomposed square wave decomposition curve, including: Based on the determined amount of sample point data, a corresponding set of Walsh basis functions is selected; Each Walsh basis function is multiplied by the coefficient vector obtained by Walsh transformation to obtain a separate decomposition curve as the decomposed square wave decomposition curve.

11. The method according to claim 2, characterized in that If the currently matched power curve decomposition algorithm is a sine sign function decomposition algorithm, the original power curve is decomposed based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve, including: Decomposing the original power curve into multiple sub-curves based on period differences, each sub-curve having a different period; Traverse each sub-curve and assign the function value of each point on each sub-curve based on the positive and negative relationship; Based on the value assigned to each point, the corresponding sub-curves are transformed, and the obtained transformed curve is used as a corresponding square wave decomposition curve.

12. The method according to claim 11, characterized in that The traversing each sub-curve and assigning the function value of each point on each sub-curve based on the positive and negative relationship includes: The points where the function value is positive are assigned the first assignment, and the points where the function value is copied are assigned the second assignment; wherein the absolute value of the first assignment is equal to the absolute value of the second assignment; The first assigned value is a positive value, and the second assigned value is a negative value.

13. The method according to claim 2, if the currently matched power curve decomposition algorithm is a combined decomposition algorithm, decomposing the original power curve based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve, comprising: Decomposing the original power curve into multiple sub-curves based on period differences, each sub-curve having a different period; Each sub-curve is traversed, and the sub-curve segment in the static power stage is assigned a value of 0, and the function value of each point on the sub-curve segment in the non-static power stage is assigned based on the positive and negative relationship.

14. The method according to claim 13, characterized in that The assigning of function values ​​of each point on the sub-curve segment of the non-stationary power stage based on a positive and negative relationship includes: Assign the points where the function value is positive as the third assignment, and assign the points where the function value is copied as the fourth assignment; wherein the absolute value of the third assignment is equal to the absolute value of the fourth assignment; The third assigned value is a positive value, and the fourth assigned value is a negative value.

15. A power curve decomposition system for an energy storage system, characterized in that: The system comprises: A collection unit, used to collect the original power curve and determine the working type of the target energy storage system based on the original power curve; A matching unit, used for matching a corresponding power curve decomposition algorithm based on a corresponding working type; The decomposition unit is used to decompose the original power curve based on the matched power curve decomposition algorithm to obtain a square wave decomposition curve corresponding to the original power curve.

16. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the energy storage system power curve decomposition method according to any one of claims 1 to 14.

Images