Multi-time-scale electric quantity balancing method and system based on multi-element energy storage complementary cooperation

Abstract

The invention discloses a multi-time scale electric quantity balancing method and system based on multi-element energy storage complementary cooperation, and relates to the technical field of electric quantity balancing, and the method comprises the following steps: obtaining a dynamic response parameter set of a multi-element energy storage medium, and constructing a response vector through clustering analysis; adopting a DTW algorithm to DTW a path distance matrix, and extracting a phase difference sequence of each energy storage medium through Hilbert transformation to obtain a candidate path set; initializing a redundancy capacity allocation vector, and randomly superposing the redundancy capacity allocation vector to the candidate path set to obtain a first evaluation matrix; constructing a multi-objective optimization function, evaluating the deviation between the first evaluation matrix and a preset electric quantity balance curve by adopting an NSGA-III algorithm, and correcting a redundant capacity allocation vector to obtain an optimal capacity allocation value sequence of each energy storage medium on a target time scale set; the problem that the scheduling efficiency and the system stability are reduced due to the fact that a traditional energy storage scheduling method does not fully consider the complex relation between the electric quantity balance target and the energy storage path dynamic response is solved.

Description

Technical Field

[0001] The present invention relates to the field of power balancing technology, and more specifically, to a multi-time-scale power balancing method and system based on complementary and coordinated multi-element energy storage. Background Art

[0002] With the large-scale integration of renewable energy and the increasing demand for flexible power system regulation, energy storage technology, as a key means of improving grid stability and operational efficiency, is gaining increasing attention. Various types of energy storage media (such as lithium batteries, supercapacitors, compressed air energy storage, and flywheel energy storage) exhibit significant differences in physical properties such as response speed, capacity density, and life cycle, making them naturally complementary. In practical applications, achieving efficient coordination between different energy storage media has become a key technical issue in the scheduling and optimization of multi-storage energy systems.

[0003] For example, the invention patent announcement with announcement number: CN117595254A discloses a multi-time-scale power and electricity balance method applied to power systems. The multi-time-scale power and electricity balance method includes a power and electricity balance optimization model and a power and electricity balance optimization process. The beneficial effect of the present invention is that it can clear technical obstacles for the smooth planning and construction of new power systems and lay a technical foundation. It analyzes the impact of different flexible power sources and adjustable loads and demand-side response changes and typical fluctuation scenarios on the traditional power and electricity balance measurement results, and finds out problems and loopholes in traditional methods.

[0004] The above disclosed technical solutions have at least the following technical problems: Currently, traditional energy storage scheduling methods often use a single time scale or ignore the dynamic response characteristics of different energy storage media. This makes it difficult to fully exploit the compensation capabilities of multi-element energy storage at different time scales (such as milliseconds, seconds, minutes, and hours). Furthermore, existing methods generally employ fixed strategies or simple heuristic algorithms for capacity allocation, failing to fully consider the complex relationship between energy balance objectives and the dynamic response of energy storage paths. This makes it difficult to balance scheduling efficiency and system stability. To address these issues, the present invention proposes a solution. Summary of the Invention

[0005] To overcome the above-mentioned shortcomings of the prior art, embodiments of the present invention provide a multi-time-scale power balancing method and system based on the complementary and coordinated development of multiple energy storage media. By integrating the dynamic characteristics of multiple energy storage media, an optimal capacity allocation value sequence of the energy storage media on a target time-scale set is obtained. This addresses the problem that traditional energy storage scheduling methods do not fully consider the complex relationship between power balancing objectives and the dynamic response of energy storage paths, resulting in reduced scheduling efficiency and system stability.

[0006] To achieve the above object, the present invention provides the following technical solutions: A multi-time-scale power balancing method based on the complementary and coordinated multi-element energy storage system includes the following steps: obtaining a dynamic response parameter set of a multi-element energy storage medium and constructing a response vector through cluster analysis; using the DTW algorithm to perform trajectory matching on the response vector to obtain a DTW path distance matrix, and extracting a phase difference sequence of each energy storage medium through Hilbert transform to obtain a candidate path set; initializing a redundant capacity allocation vector and randomly superimposing it on the candidate path set to obtain a first evaluation matrix; constructing a multi-objective optimization function and using the NSGA-III algorithm to evaluate the deviation between the first evaluation matrix and a preset power balance curve; and correcting the redundant capacity allocation vector based on the evaluation result to obtain an optimal capacity allocation value sequence for each energy storage medium on the target time-scale set.

[0007] In a preferred embodiment, the constructing of the response vector by cluster analysis specifically comprises: performing feature analysis on the dynamic response parameter set using a principal component analysis method to obtain a plurality of first response vectors; grouping the plurality of first response vectors according to a preset energy storage medium type, and calculating the Euclidean distance of the first response vectors in each group to obtain a plurality of response distance clusters; performing cluster analysis on the plurality of response distance clusters using a density clustering algorithm to obtain a cluster center vector of each energy storage medium; and performing weighted fusion on the cluster center vectors of each energy storage medium according to a preset weight to obtain a response vector.

[0008] In a preferred embodiment, the DTW algorithm is used to perform trajectory matching on the response vector to obtain a DTW path distance matrix, and the phase difference sequence of each energy storage medium is extracted through Hilbert transform to obtain a candidate path set. Specifically, the following steps are performed: obtaining a historical standard response trajectory, and using the DTW algorithm to calculate the matching path between the response vector and the historical standard response trajectory to obtain several path matching sequences; based on the several path matching sequences, a DTW path distance matrix is ​​constructed; performing frequency domain decomposition on the DTW path distance matrix to extract the instantaneous phase characteristics of each energy storage medium; converting the instantaneous phase characteristics into a phase difference sequence through Hilbert transform; and screening the several path matching sequences according to the phase difference sequence to generate a candidate path set.

[0009] In a preferred embodiment, the initial redundancy capacity allocation vector is randomly generated in the feasible solution space, specifically: a clustering algorithm is used to partition the feasible solution space to obtain a plurality of capacity clustering centers; the plurality of capacity clustering centers are adjusted to perform biased random sampling on the preset probability density function to obtain the initial redundancy capacity allocation vector.

[0010] In a preferred embodiment, the initial redundancy capacity allocation vector is randomly generated in the feasible solution space, specifically: a clustering algorithm is used to partition the feasible solution space to obtain a plurality of capacity clustering centers; the plurality of capacity clustering centers are adjusted to perform biased random sampling on the preset probability density function to obtain the initial redundancy capacity allocation vector.

[0011] In a preferred embodiment, the first evaluation matrix is constructed by vector superposition operation, specifically: the redundancy capacity allocation vector is decomposed by wavelet to obtain a basic capacity component and a fluctuation compensation component; the basic capacity component is linearly superimposed on the reference capacity of the candidate path set; the fluctuation compensation component is fused with the phase difference sequence of the candidate path by convolution operation to generate an adjusted capacity; the reference capacity and the adjusted capacity are fused according to a preset weight to obtain the first evaluation matrix.

[0012] In a preferred embodiment, the deviation of the first evaluation matrix from the preset power balance curve is evaluated by the NSGA-III algorithm, specifically: the first evaluation matrix is input into the NSGA-III algorithm, a reference point set and a population are initialized, and non-dominated sorting is performed to obtain a first candidate scheme set; the power output sequence corresponding to each candidate path in the first evaluation matrix is extracted and compared with the preset power balance curve to construct a first deviation vector; the first candidate scheme set is screened based on the first deviation vector to obtain a second candidate scheme set.

[0013] In a preferred embodiment, the redundancy capacity allocation vector is corrected according to the evaluation result to obtain a best capacity allocation value sequence of each energy storage medium on the target time scale set, specifically: The redundancy vector component corresponding to the Pareto solution is extracted from the second candidate scheme set; the gradient direction deviation of each redundancy vector component from the preset initial component is calculated to obtain a correction step; the redundancy capacity allocation vector is corrected based on the correction step to obtain a corrected vector; the corrected vector is analyzed on the target time scale set to obtain a plurality of time scale components; the plurality of time scale components are respectively converted into actual capacity instruction sequences and injected into the energy storage control system interface protocol.

[0014] The technical effects and advantages of the multi-time-scale power balancing method and system based on the complementary and coordinated multi-element energy storage of the present invention are as follows: 1. This invention performs clustering and principal component analysis on the dynamic response parameters of multiple energy storage media to extract representative response vectors, effectively reflecting the differences in the response capabilities of various energy storage media at different time scales. It then combines the DTW dynamic time warping algorithm with the Hilbert transform to extract phase difference features, thereby establishing a set of candidate paths and accurately modeling the dynamic characteristics of the energy storage paths. Furthermore, wavelet decomposition and convolution fusion are introduced to superimpose the redundant capacity allocation vectors with the dynamic characteristics of the energy storage paths, forming a multi-time-scale evaluation matrix, providing refined data support for subsequent multi-objective optimization. The NSGA-III algorithm is used to perform non-dominated sorting and deviation analysis on the capacity allocation schemes, further improving the rationality of the redundant allocation and the coordination of the system output. Finally, the redundant allocation vectors are iteratively corrected based on the multi-objective optimization results to form a time-scaled capacity instruction sequence, enabling refined capacity allocation under dynamic load changes. This method significantly improves the response efficiency and operational stability of multiple energy storage systems in complex power grid environments, addressing the problems of crude scheduling strategies and lack of dynamic coupling analysis in traditional methods, and providing technical support and implementation path for building energy storage optimization scheduling systems for new smart grids. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Schematic diagram of the process of the multi-time scale power balancing method based on the complementary cooperation of multiple energy storages of the present invention; Figure 2 A schematic diagram of the decomposition results of the initial redundant capacity allocation vector and the modified vector on a time scale provided by an embodiment of the present invention; Figure 3 This is a structural diagram of the multi-time-scale power balancing system based on the complementary and coordinated multi-element energy storage system of the present invention. DETAILED DESCRIPTION

[0016] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0017] Example 1, Figure 1 The present invention provides a multi-time scale power balancing method based on the complementary cooperation of multiple energy storages, including the following steps: S1, obtain the dynamic response parameter set of the multivariate energy storage medium and construct the response vector through cluster analysis; In this example, the dynamic response parameter set of the multi-element energy storage medium is obtained, specifically: First, modeling experiments were conducted on three typical types of grid-connected energy storage media: lithium batteries, supercapacitors, and compressed air energy storage devices. Dynamic response data was obtained under standard test conditions (ambient temperature 25°C, rated load fluctuation ±10%, and a cycle of 30 seconds). The experiments included unit power step input tests, load disturbance response tests, and power frequency sweep tests.

[0018] During the experiment, a system identification method was used to extract dynamic parameters. Specifically, the step response test was used to calculate the response delay (the time point when the output power first exceeds the target value by 5%) and the dynamic regulation rate (the maximum slope of the power rise curve divided by the rated capacity, that is, the rate of change in C / s). The load disturbance test was used to calculate the power tracking error using the root mean square error formula, as follows:

[0019] Where N is the total number of sampling points; is the actual output power measured at the kth sampling point; is the power reference command value at the kth sampling point; is the rated power of the energy storage medium.

[0020] The experimental results are: For lithium batteries: the response delay is 0.5s, the dynamic regulation rate is 0.2C / s, the maximum charge and discharge power is 100kW, and the RMS power tracking error is 3.1%; For supercapacitors: response delay is 0.1s, dynamic adjustment rate is 1.0C / s, maximum output power is 60kW, tracking error is 1.2%, and inertial response time constant is 0.03s; For the compressed air energy storage system: the response delay is 2.5s, the regulation rate is 0.05C / s, the maximum output power is 300kW, the tracking error is 5.6%, and the heat exchange delay time is 4.1s; The above parameters are normalized to form a dynamic response parameter set. The normalization method uses the fixed interval Min–Max method. The relevant formula is as follows:

[0021] Where x is the original parameter value to be normalized; The engineering lower limit value set for this parameter; The engineering upper limit value set for this parameter; It is a dimensionless result after normalization, and its value ranges from 0 to 1.

[0022] To ensure reproducibility, this embodiment adopts unified engineering upper and lower limits. For example, the upper limit of response delay is set to 10s, the upper limit of adjustment rate is set to 2C / s, the upper limit of power error is set to 10%, and the upper limit of time constant is set to 100s. Values ​​outside the range are truncated according to the boundaries.

[0023] In this example, a response vector is constructed through cluster analysis, specifically: The principal component analysis method is used to perform characteristic analysis on the dynamic response parameter set to obtain several first response vectors; Grouping a plurality of first response vectors according to a preset energy storage medium type, and calculating the Euclidean distance of the first response vector in each group to obtain a plurality of response distance clusters; The density clustering algorithm is used to perform cluster analysis on several response distance clusters to obtain the cluster center vector of each energy storage medium; The cluster center vectors of each energy storage medium are weightedly fused according to the preset weights to obtain the response vector.

[0024] The above steps are implemented as follows: Firstly, the principal component analysis (PCA) method is used to reduce the dimension and extract the features of the dynamic response parameter set. Specifically, the parameter matrix After standardization, the covariance matrix is ​​calculated and its eigenvalues ​​and eigenvectors are obtained. The eigenvalues ​​are sorted from large to small, and the first several principal components with a cumulative contribution rate of ≥ 95% are selected to form the first response vector set. .

[0025] Subsequently, these first response vectors are grouped and classified according to the energy storage medium type (such as lithium battery, supercapacitor, compressed air, etc.), and the Euclidean distance between each response vector in each group is calculated. The calculation formula is:

[0026] in, 、 Represent the two first response vectors, 、 express 、 The kth component of , m is the dimension of the principal component. Response distance clusters are formed that reflect the internal response differences of the energy storage medium.

[0027] On this basis, a density clustering algorithm (such as DBSCAN) is used to perform cluster analysis on the response distance clusters. The neighborhood radius parameter ε can be adaptively set according to the mean value and standard deviation of the response distance (for example, ε = mean value + 0.5 × standard deviation). The minimum number of samples is 3 to ensure the robustness of the clustering and identify the representative response feature points in each type of energy storage medium, namely the cluster center vector.

[0028] Finally, the center vectors of each cluster are weighted fused according to the preset weights, and the calculation formula is:

[0029] in, is the cluster center vector of the i-th energy storage medium, is the corresponding weight, satisfying , q is the number of energy storage medium types. The weight can be set according to capacity ratio, control priority or empirical parameters. The final response vector is generated , serving as the foundational data for subsequent path matching and capacity optimization analysis. This approach fully accounts for the dynamic response differences of different types of energy storage devices and uniformly represents them in a structured, characterized form, improving overall modeling accuracy and the practicality of dispatch control.

[0030] S2, using the DTW algorithm to perform trajectory matching on the response vector, obtaining the DTW path distance matrix, and extracting the phase difference sequence of each energy storage medium through Hilbert transform to obtain the candidate path set; In this example, the DTW algorithm is used to perform trajectory matching on the response vector to obtain the DTW path distance matrix. The phase difference sequence of each energy storage medium is extracted through Hilbert transform to obtain the candidate path set, which is specifically: Obtain historical standard response trajectories, and use the DTW algorithm to calculate the matching paths between the response vector and the historical standard response trajectories to obtain several path matching sequences; Based on several path matching sequences, a DTW path distance matrix is ​​constructed; Perform frequency domain decomposition on the DTW path distance matrix to extract the instantaneous phase characteristics of each energy storage medium; The instantaneous phase feature is converted into a phase difference sequence through Hilbert transform; Several path matching sequences are screened according to the phase difference sequence to generate a candidate path set.

[0031] The above steps are implemented as follows: Calculate the response vector using the dynamic time warping (DTW) algorithm Response trajectory with historical standards The distance measurement formula for the matching path between is:

[0032] And by the recurrence relation:

[0033] Find the optimal alignment path, where is the local distance between the i-th and j-th sampling points.

[0034] After obtaining multiple path matching sequences, construct the DTW path distance matrix , where the elements It represents the average DTW distance between the p-th path and the q-th path, and is used to quantify the matching degree of different paths.

[0035] Pair Matrix Each path sequence in the frequency domain is decomposed using fast Fourier transform (FFT) to obtain the amplitude sequence of different frequency components. Then the instantaneous phase is calculated using the Hilbert transform, and the calculation formula is as follows:

[0036] in, represents the Hilbert transform operator, is the time domain signal amplitude.

[0037] According to the difference in instantaneous phase of different energy storage media, the phase difference sequence is defined as:

[0038] in, and are the instantaneous phases of energy storage media i and j respectively.

[0039] Finally, the path screening is performed based on the stability index of the phase difference sequence, which is defined as the standard deviation of the phase difference sequence ,when rad, it is judged as "trend stable"; at the same time, the matching degree between the path and the standard trajectory is calculated ,when When , it is judged as "high degree of coordination". The paths that meet the above conditions constitute the candidate path set. Among them, The unit is radian (rad), which represents the standard deviation of the phase difference sequence; Represents the dynamic time warping cumulative distance between a candidate path and the standard response trajectory.

[0040] S3, initialize the redundant capacity allocation vector and randomly superimpose it on the candidate path set to obtain the first evaluation matrix; In this example, the redundant capacity allocation vector is initialized and randomly superimposed on the candidate path set to obtain the first evaluation matrix, which is specifically: Construct a feasible solution space for redundant capacity allocation vectors based on the preset grid load fluctuation range and energy storage medium characteristics; Randomly generate an initial redundant capacity allocation vector in the feasible solution space; Mapping the initial redundant capacity allocation vector to each candidate path in the candidate path set according to the energy storage medium type to obtain a redundant capacity allocation vector for each candidate path; Based on the redundant capacity allocation vector of each candidate path, a first evaluation matrix is ​​constructed through vector superposition operation.

[0041] The above steps are implemented as follows: First, based on the load fluctuation range of the power grid system and the response characteristics of various energy storage media, a feasible solution space for redundant capacity allocation is established. The feasible solution space is defined by constraints, including: the SOC range of the energy storage medium , power range , and charge and discharge efficiency , ensuring that the randomly generated vectors meet energy conservation and equipment safety boundaries. This space covers all possible redundant capacity configurations that are technically feasible and meet safe operation boundaries.

[0042] On this basis, the method randomly generates an initial redundant capacity allocation vector in the solution space, which represents the redundant capacity shared by different energy storage media under the current load fluctuation background.

[0043] In this example, the initial redundant capacity allocation vector is randomly generated in the feasible solution space, specifically: Use clustering algorithms (such as K-means or DBSCAN) to partition the feasible solution space and obtain several capacity cluster centers; Adjust several capacity cluster centers to perform biased random sampling corresponding to the preset probability density function to obtain the initial redundant capacity allocation vector, such as Gaussian distribution:

[0044] in, is the cluster center, is the standard deviation, x represents the capacity allocation vector, which is a value sampled from the probability distribution and represents the redundant capacity allocation of each energy storage medium; or Beta distribution:

[0045] in, , is the shape parameter, This mechanism effectively avoids the problems of uneven distribution or falling into infeasible areas that may be caused by purely random generation, and provides a stable and accurate initial input basis for subsequent path superposition and optimization.

[0046] In this example, the first evaluation matrix is ​​constructed by vector superposition operation, specifically: Perform wavelet decomposition on the redundant capacity allocation vector to obtain the basic capacity component and the fluctuation compensation component; Linearly superimpose the basic capacity component to the baseline capacity of the candidate path set; The fluctuation compensation component is fused with the phase difference sequence of the candidate path through convolution operation to generate the adjustment capacity; The baseline capacity and the adjusted capacity are integrated according to preset weights to obtain a first evaluation matrix.

[0047] It should be noted that: Wavelet decomposition uses the Daubechies (db4) wavelet basis for three-layer decomposition, with the low-frequency part as the basic capacity component and the high-frequency part as the fluctuation compensation component. The convolution operation uses the formula:

[0048] in, To adjust the capacity, is the fluctuation compensation component, is the phase difference sequence, is the time delay parameter, which indicates the time offset of the fluctuation compensation component. The weighted fusion adopts the formula:

[0049] in, is the capacity sequence in the first evaluation matrix, is the base capacity, It is the fusion weight, ranging from 0 to 1, and can be set according to the system scheduling strategy.

[0050] This evaluation matrix not only includes the time-series output characteristics of various energy storage media, but also comprehensively considers factors such as power fluctuation, phase response, and path adaptation. It provides a structured and quantifiable input basis for the subsequent use of multi-objective optimization algorithms (such as NSGA-III) for capacity configuration and scheduling path selection, thereby significantly improving the balance performance and operating efficiency of the energy storage system at multiple time scales.

[0051] S4, constructing a multi-objective optimization function and using the NSGA-III algorithm to evaluate the deviation between the first evaluation matrix and the preset power balance curve; In this example, the NSGA-III algorithm is used to evaluate the deviation between the first evaluation matrix and the preset power balance curve, specifically: Input the first evaluation matrix into the NSGA-III algorithm, initialize the reference point set and population, and perform non-dominated sorting to obtain the first candidate solution set; Extracting a power output sequence corresponding to each candidate path in the first evaluation matrix and comparing it with a preset power balance curve to construct a first deviation vector; The first candidate solution set is screened based on the first deviation vector to obtain a second candidate solution set.

[0052] It should be noted that the first evaluation matrix is ​​first imported into the NSGA-III optimization framework as input data, and the reference point set and population individuals are initialized. The population size is set to 100, and the maximum number of iterations is set to 200 generations. The reference point set is generated by uniformly distributing it on the target dimension. Subsequently, all evaluation solutions are classified by the non-dominated sorting method to form a preliminary set of candidate solutions.

[0053] In terms of optimization objectives, the multi-objective function is defined as:

[0054] in,

[0055] represents the mean absolute error between the path power output and the power balance curve, is the actual output power of the candidate path at time t; is the power reference command value of the preset power balance curve at time t, and T is the total duration of the optimization cycle;

[0056] represents the capacity utilization, where is the energy actually used during the optimization period, is the total available energy;

[0057] It represents the standard deviation of the path output power series and is used to measure the stationarity of the path.

[0058] Next, the corresponding power output time series is extracted from each candidate path and compared point by point with the pre-set power balance curve. The difference between the two is calculated to construct a "first deviation vector" that reflects the degree of power deviation. The deviation vector is defined as:

[0059] This deviation vector is used to measure each candidate path's performance in meeting the power balance objective and serves as an evaluation metric for the optimization algorithm. Based on this metric, the algorithm further selects path combinations with superior performance and, through the Pareto front, selects a second set of candidate solutions, representing the optimal solution set that achieves a good trade-off between multiple objectives (such as capacity utilization, path smoothness, and minimizing power deviation). Through this optimization and evaluation process, the method effectively identifies path combinations with the best responsiveness and power allocation performance at different time scales, thereby improving the scheduling accuracy and global coordination of the energy storage system.

[0060] S5, the redundant capacity allocation vector is modified according to the evaluation results to obtain the optimal capacity allocation value sequence of each energy storage medium on the target time scale set.

[0061] In this example, the redundant capacity allocation vector is modified according to the evaluation results to obtain the optimal capacity allocation value sequence of each energy storage medium on the target time scale set, specifically: Extracting redundant vector components corresponding to the Pareto solution from the second candidate solution set; Calculate the gradient direction deviation between each redundant vector component and the preset initial component to obtain the correction step size; Correcting the redundant capacity allocation vector based on the correction step size to obtain a correction vector; The correction vector is analyzed on the target time scale set to obtain several time scale components; Several time scale components are converted into actual capacity instruction sequences and injected into the energy storage control system interface protocol.

[0062] For example, let the second candidate solution set be:

[0063] in, is the kth candidate solution, corresponding to the redundant capacity allocation vector of different energy storage media; N is the number of candidate solutions; m is the total number of energy storage media; represents the redundant capacity allocation value of the j-th energy storage medium in the k-th solution.

[0064] The gradient direction deviation between the current component and the original component in each dimension is defined as:

[0065] in, Represents the capacity gradient direction deviation of the j-th energy storage medium; is the redundant capacity allocation value corresponding to the j-th energy storage medium in the Pareto optimal solution; Represents the initial redundant capacity allocation value of the j-th energy storage medium; j is an index variable, indicating the energy storage medium number.

[0066] Based on the gradient direction deviation, the correction step size is defined as:

[0067] in, is the correction step size of the j-th energy storage medium, is the learning rate parameter (range 0 to 1, used to control the correction amplitude).

[0068] The corrected capacity vector is:

[0069] in, is the revised redundant capacity allocation value of the j-th energy storage medium.

[0070] It should be noted that the redundant capacity vector corresponding to the Pareto optimal solution is first extracted from the second candidate solution set and compared with the initial component initially set, and its gradient direction deviation in the multidimensional parameter space is calculated, thereby determining the correction step size of each energy storage medium in the capacity adjustment process. Subsequently, the correction step size is used to make a directional adjustment to the original redundant capacity allocation vector to generate a new correction vector. This vector is then mapped to the target time scale set, and several capacity components of different frequencies and durations are extracted by time scale decomposition, such as Figure 2 The figure shows the decomposition results of the initial redundant capacity allocation vector and the corrected vector on the time scale, where the low-frequency component reflects the long-term capacity configuration and the high-frequency component corresponds to the compensation of short-term fluctuations. Each time scale component is further converted into a capacity control instruction sequence with practical engineering significance, and finally injected into the dispatching system through the control interface of the energy storage system, realizing closed-loop control from theoretical optimization to actual execution. This process not only improves the accuracy and timeliness of capacity allocation, but also ensures the system's dynamic response capability to power fluctuations at multiple time scales, effectively solving the problems of rigid capacity configuration and difficulty in adapting to dynamic changes in the power system in traditional methods.

[0071] Example 2, Figure 3 The present invention provides a multi-time-scale power balancing system based on the complementary cooperation of multiple energy storages, including a vector construction module, a phase matching module, a redundancy assessment module, an optimization decision module, and a capacity allocation module: Vector construction module, used to obtain the dynamic response parameter set of the multivariate energy storage medium and construct the response vector through cluster analysis; The phase matching module is used to perform trajectory matching on the response vector using the DTW algorithm to obtain the DTW path distance matrix, and extract the phase difference sequence of each energy storage medium through Hilbert transform to obtain the candidate path set; A redundancy evaluation module is used to initialize a redundant capacity allocation vector and randomly superimpose it on the candidate path set to obtain a first evaluation matrix; An optimization decision module is used to construct a multi-objective optimization function and use the NSGA-III algorithm to evaluate the deviation between the first evaluation matrix and the preset power balance curve; The capacity allocation module is used to modify the redundant capacity allocation vector according to the evaluation results to obtain the optimal capacity allocation value sequence of each energy storage medium on the target time scale set.

[0072] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0073] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0074] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0075] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0076] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0077] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-time scale electricity balancing method based on the complementary cooperation of multiple energy storage systems, characterized by: The following steps are involved: Obtain the dynamic response parameter set of the multivariate energy storage medium and construct the response vector through cluster analysis; The DTW algorithm is used to perform trajectory matching on the response vector to obtain the DTW path distance matrix, and the phase difference sequence of each energy storage medium is extracted through Hilbert transform to obtain the candidate path set. Initialize the redundant capacity allocation vector and randomly superimpose it on the candidate path set to obtain the first evaluation matrix; Construct a multi-objective optimization function and use the NSGA-III algorithm to evaluate the deviation between the first evaluation matrix and the preset power balance curve; The redundant capacity allocation vector is modified according to the evaluation results to obtain the optimal capacity allocation value sequence of each energy storage medium on the target time scale set.

2. The multi-time scale power balancing method based on multi-element energy storage complementary synergy according to claim 1 is characterized in that: The response vector is constructed by cluster analysis, specifically: The principal component analysis method is used to perform characteristic analysis on the dynamic response parameter set to obtain several first response vectors; Grouping a plurality of first response vectors according to a preset energy storage medium type, and calculating the Euclidean distance of the first response vector in each group to obtain a plurality of response distance clusters; The density clustering algorithm is used to perform cluster analysis on several response distance clusters to obtain the cluster center vector of each energy storage medium; The cluster center vectors of each energy storage medium are weightedly fused according to the preset weights to obtain the response vector.

3. The multi-time scale power balancing method based on multi-element energy storage complementary synergy according to claim 2 is characterized in that: The DTW algorithm is used to perform trajectory matching on the response vector to obtain a DTW path distance matrix, and the phase difference sequence of each energy storage medium is extracted through Hilbert transform to obtain a candidate path set, specifically: Obtain historical standard response trajectories, and use the DTW algorithm to calculate the matching paths between the response vector and the historical standard response trajectories to obtain several path matching sequences; Based on several path matching sequences, a DTW path distance matrix is ​​constructed; Perform frequency domain decomposition on the DTW path distance matrix to extract the instantaneous phase characteristics of each energy storage medium; The instantaneous phase feature is converted into a phase difference sequence through Hilbert transform; Several path matching sequences are screened according to the phase difference sequence to generate a candidate path set.

4. The multi-time scale power balancing method based on multi-element energy storage complementary synergy according to claim 3 is characterized in that: The redundant capacity allocation vector is initialized and randomly superimposed on the candidate path set to obtain a first evaluation matrix, specifically: Construct a feasible solution space for redundant capacity allocation vectors based on the preset grid load fluctuation range and energy storage medium characteristics; Randomly generate an initial redundant capacity allocation vector in the feasible solution space; Mapping the initial redundant capacity allocation vector to each candidate path in the candidate path set according to the energy storage medium type to obtain a redundant capacity allocation vector for each candidate path; Based on the redundant capacity allocation vector of each candidate path, a first evaluation matrix is ​​constructed through vector superposition operation.

5. The multi-time-scale power balancing method based on multi-element energy storage complementary synergy according to claim 4 is characterized in that: The initial redundant capacity allocation vector is randomly generated in the feasible solution space, specifically: Clustering algorithm is used to partition the feasible solution space and obtain several capacity cluster centers; Adjust several capacity cluster centers to perform biased random sampling corresponding to the preset probability density function to obtain the initial redundant capacity allocation vector.

6. The multi-time-scale power balancing method based on multi-element energy storage complementary synergy according to claim 5 is characterized in that: The first evaluation matrix is ​​constructed by vector superposition operation, specifically: Perform wavelet decomposition on the redundant capacity allocation vector to obtain the basic capacity component and the fluctuation compensation component; Linearly superimpose the basic capacity component to the baseline capacity of the candidate path set; The fluctuation compensation component is fused with the phase difference sequence of the candidate path through convolution operation to generate the adjustment capacity; The baseline capacity and the adjusted capacity are integrated according to preset weights to obtain a first evaluation matrix.

7. The multi-time-scale power balancing method based on multi-element energy storage complementary synergy according to claim 6 is characterized in that: The NSGA-III algorithm is used to evaluate the deviation between the first evaluation matrix and the preset power balance curve, specifically: Input the first evaluation matrix into the NSGA-III algorithm, initialize the reference point set and population, and perform non-dominated sorting to obtain the first candidate solution set; Extracting a power output sequence corresponding to each candidate path in the first evaluation matrix and comparing it with a preset power balance curve to construct a first deviation vector; The first candidate solution set is screened based on the first deviation vector to obtain a second candidate solution set as an evaluation result.

8. The multi-time-scale power balancing method based on multi-element energy storage complementary synergy according to claim 7 is characterized in that: The redundant capacity allocation vector is modified according to the evaluation results to obtain the optimal capacity allocation value sequence of each energy storage medium on the target time scale set, specifically: Extracting redundant vector components corresponding to the Pareto solution from the second candidate solution set; Calculate the gradient direction deviation between each redundant vector component and the preset initial component to obtain the correction step size; Correcting the redundant capacity allocation vector based on the correction step size to obtain a correction vector; The correction vector is analyzed on the target time scale set to obtain several time scale components; Several time scale components are converted into actual capacity instruction sequences and injected into the energy storage control system interface protocol.

9. A multi-time-scale power balancing system based on the complementary cooperation of multiple energy storages, applied to the multi-time-scale power balancing method based on the complementary cooperation of multiple energy storages according to any one of claims 1 to 8, characterized in that: It includes vector construction module, phase matching module, redundancy assessment module, optimization decision module and capacity allocation module: Vector construction module, used to obtain the dynamic response parameter set of the multivariate energy storage medium and construct the response vector through cluster analysis; The phase matching module is used to perform trajectory matching on the response vector using the DTW algorithm to obtain the DTW path distance matrix, and extract the phase difference sequence of each energy storage medium through Hilbert transform to obtain the candidate path set; A redundancy evaluation module is used to initialize a redundant capacity allocation vector and randomly superimpose it on the candidate path set to obtain a first evaluation matrix; An optimization decision module is used to construct a multi-objective optimization function and use the NSGA-III algorithm to evaluate the deviation between the first evaluation matrix and the preset power balance curve; The capacity allocation module is used to modify the redundant capacity allocation vector according to the evaluation results to obtain the optimal capacity allocation value sequence of each energy storage medium on the target time scale set.

Images