Energy storage layout optimization method and system for new energy station cluster

Abstract

The invention discloses a new energy station cluster-oriented energy storage layout optimization method and system. The method comprises the steps of obtaining historical output data of new energy stations; an optimal vine Copula structure is determined, and massive scenes of output of multiple new energy stations under different time sections are obtained; a novel power system typical operation scene is generated; in each typical operation scene, an energy storage economy evaluation index, a peak regulation evaluation index, a new energy output time sequence fluctuation stabilizing capability index, a new energy gradeability stabilizing capability evaluation index and a frequency active support capability evaluation index are constructed; constructing a mathematical model considering the energy storage layout optimization problem of the following network type and the construction network type based on the constructed indexes; and performing iterative solution on the mathematical model based on dynamic step size simulation and a differential evolution algorithm to obtain an optimal energy storage layout scheme. According to the obtained energy storage layout scheme, the safety and stability of the power system can be improved from multiple aspects.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage layout optimization, and particularly to an energy storage layout optimization method and system for a new energy power station cluster. Background Technique

[0002] The statements in this part only provide background technical information related to the present invention and do not necessarily constitute prior art.

[0003] With the development of energy storage technology, its cost is continuously reduced and its control performance is continuously enhanced, which plays an important role in suppressing the problems of power grid operation caused by new energy. From the perspective of the role of energy storage itself, it can provide optimization and flexibility of the power system. Through the energy storage system, the power grid load can be adjusted, the supply-demand relationship can be balanced, and the efficiency and stability of the power system can be improved; through the energy storage system, the excess power of renewable energy can be stored for use when there is a shortage, which can solve the intermittency and volatility of renewable energy such as solar energy and wind energy, and improve the utilization rate of renewable energy.

[0004] The proportion of new energy is increasing continuously, and the inherent randomness and volatility of new energy pose a huge challenge to the safety and stability of the new power system. Energy storage technology is an important means to cope with the uncertainty of new energy.

[0005] Most of the existing energy storage combination optimization layouts focus on the energy storage planning and configuration of a specific scenario, with less demonstration of the overall energy storage planning scale considering the multi-reuse function of energy storage, and at the same time, there is a lack of research on the energy storage planning layout optimization method. In particular, the existing methods do not specifically consider the new type of network-forming energy storage, and it is difficult to play its role in the safe and stable operation of the power system. Summary of the Invention

[0006] In order to solve the above problems, the present invention proposes an energy storage layout optimization method and system for a new energy power station cluster, which fully considers the spatial correlation between the outputs of new energy power stations, and at the same time comprehensively considers two different types of energy storage, namely following the grid and forming the grid, as well as the role of energy storage in frequency modulation, peak shaving and active frequency support, so as to obtain an optimal energy storage layout plan.

[0007] In some embodiments, the following technical solutions are adopted:

[0008] An energy storage layout optimization method for a new energy power station cluster, comprising:

[0009] Obtain the historical output data of new energy power stations;

[0010] Determine the optimal vine Copula's first tree structure based on the mean-shift clustering algorithm, and then determine the optimal vine Copula structure. Based on the optimal vine Copula structure, determine the optimal Copula function, sample the optimal vine Copula function and perform inverse transformation to obtain a large number of scenarios of the output of multiple new energy power stations at different time sections;

[0011] Reduce the large number of scenarios of the output of new energy power stations at each time section to obtain typical scenarios of new energy output, and generate typical operating scenarios of the new power system by splicing in the time dimension;

[0012] Under each typical operating scenario, construct an energy storage economic evaluation index, a peak shaving evaluation index, an index for the ability to suppress the time-series volatility of new energy output, an index for the ability to suppress the ramping of new energy, and an index for the ability to actively support the frequency; construct a mathematical model for the energy storage layout optimization problem considering grid-connected and network-forming types based on the constructed indexes;

[0013] Based on the dynamic step-size simulation and differential evolution algorithm, iteratively solve the mathematical model to obtain the optimal energy storage layout plan.

[0014] As an optional solution, the process of determining the optimal vine Copula structure is as follows:

[0015] Assume there are m new energy power stations in total, and calculate the Pearson correlation coefficient matrix of the output between the power stations;

[0016] Define the spatial correlation coordinate \(x_i\) of the \(i\)-th new energy power station i as the \(i\)-th row of the correlation coefficient matrix;

[0017] Cluster the new energy power stations based on the spatial correlation coordinates through the mean-shift clustering algorithm;

[0018] Suppose the number of clustering categories after clustering is n. For each power station in each category, connect them using the D-vine structure. Then, select any two power stations in different categories to connect the vine structure to realize the selection of the optimal vine Copula tree 1 structure; for the subsequent tree structures, use the D-vine structure to obtain the optimal vine Copula structure.

[0019] As an optional solution, the process of determining the optimal Copula function based on the optimal vine Copula structure is as follows:

[0020] Determine the empirical distribution function of the output of each new energy power station, and determine the empirical Copula function based on the empirical distribution function;

[0021] Each edge in the optimal vine Copula structure corresponds to a Copula function, and the parameters of each Copula function are obtained based on the maximum likelihood estimation;

[0022] Calculate the Euclidean distance between each Copula function corresponding to the current edge and the empirical Copula function, and select the Copula function with the smallest Euclidean distance as the optimal Copula function for this edge;

[0023] Determine the optimal Copula function corresponding to each edge of the optimal vine Copula structure in turn, so as to obtain the optimal vine Copula function.

[0024] As an alternative solution, simplify the massive scenarios of the new energy power station output under each time section. The specific process is as follows:

[0025] (1) Initialize N r ;

[0026] (2) Traverse any N r scenarios as the initial center points, and use the N r scenarios with the best clustering results as the initial clustering center points of the K-means clustering algorithm; randomly select a scenario from the remaining scenarios and N r initial clustering center points as the new initial clustering center points for clustering; select the N r +1 samples with the best clustering results as the initial points of the K-means clustering algorithm;

[0027] (3) Judge whether N r reaches the number of typical scenarios of new energy output. If it reaches, output the clustering result. Otherwise, let N r = N r +1, repeat the process of (2) until N r reaches the number of typical scenarios of new energy output, and output the clustering result as the typical scenarios of new energy output under each time section.

[0028] As an alternative solution, construct the energy storage economy evaluation index C EES , specifically:

[0029]

[0030] Among them, N EES is the number of energy storage power stations, c 1 is the unit capacity cost of the grid-connected energy storage, and c 2 is the unit power cost of the grid-forming energy storage; E i E1 , E i E2 are the capacities of the grid-connected and grid-forming energy storages configured for the i-th new energy power station respectively;

[0031] Or, construct the peak shaving evaluation index Y peek, specifically as follows:

[0032]

[0033] Among them, I [条件] is an indicator function, which takes the value of 1 if the condition is satisfied, and 0 otherwise; t is the time length of the time series scenario, and L 1 (t) represents the net load power curve of the system when the energy storage does not participate in the system peak shaving, and L 2 (t) represents the net load power curve of the system after the energy storage participates in peak shaving, and P 1 represents the peak shaving boundary, and P 2 represents the valley filling boundary.

[0034] Alternatively, construct an index Y for suppressing the time series volatility of new energy output 1 , specifically as follows:

[0035]

[0036] Among them, P i total (j) is the total tie-line power of the i-th energy storage station at the j-th moment, and P base,i is the reported reference capacity;

[0037] Alternatively, construct an evaluation index for suppressing the ramp rate of new energy, specifically as follows:

[0038]

[0039] Among them, Y i ramp (x) is the output ramp rate of the i-th new energy station at time x, Δt is the time interval, and P i total (x) is the total tie-line power of the i-th new energy station at time x;

[0040] Alternatively, construct an evaluation index for active frequency support ability, specifically as follows:

[0041]

[0042] Among them, t 1 is the moment when the active power deficit of the system occurs, and t 2 is the moment when the primary frequency regulation of the system ends, and f 1 (t) represents the frequency response curve of the system after the energy storage station participates in active frequency support, and f 2 (t) represents the frequency response curve of the system when the energy storage station does not participate in active frequency support.

[0043] As an alternative solution, based on the constructed index, construct a mathematical model for the energy storage layout optimization problem considering grid-connected and network-forming energy storage, specifically as follows:

[0044]

[0045] Among them, c 1 is the cost per unit capacity of the grid-following energy storage unit, and c 2 is the cost per unit power of the grid-forming energy storage unit; E i E1 and E i E2 are the capacities of the grid-following and grid-forming energy storage units configured for the i-th new energy power station respectively; m is the number of new energy power stations, and Y peek_set is the defined lower limit of the energy storage peak shaving index, and P i wind (t) is the active power of the wind power of the i-th new energy power station at time t, and P i pv (t) is the active power of the photovoltaic power of the i-th new energy power station at time t, and p i E1 (t) is the active power of the grid-following energy storage of the i-th new energy power station at time t, and p i E2 (t) is the active power of the grid-forming energy storage of the i-th new energy power station at time t, and P base,i is the reported reference capacity, and δ is the maximum allowable boundary of the tie-line power fluctuation, where δ > 0, and Y i ramp_l and Y i ramp_h are the minimum and maximum values of the new energy ramp rate of the i-th new energy power station respectively, and d f is the minimum frequency active support ability boundary of the energy storage; P i E1 is the power upper limit of the grid-following energy storage of the i-th new energy power station, and P i E2 is the power upper limit of the grid-forming energy storage of the i-th new energy power station; SOC i 1 (t) is the state of charge of the grid-following energy storage of the i-th new energy power station at time t, and SOC i 1m (t) is the minimum state of charge of the grid-following energy storage of the i-th new energy power station, and SOC i 1M (t) is the maximum state of charge of the grid-following energy storage of the i-th new energy power station, and SOC i 2 (t) is the state of charge of the grid-forming energy storage of the i-th new energy power station at time t, and SOC i 2m (t) is the minimum state of charge of the grid-forming energy storage of the i-th new energy power station, and SOC i 2M(t) is the maximum state of charge of the grid-forming energy storage for the i-th new energy power station.

[0046] As an alternative solution, based on dynamic step-size simulation and differential evolution algorithm, the mathematical model is iteratively solved to obtain the optimal energy storage layout plan, specifically:

[0047] Initialize the energy storage layout plan, and use dynamic step-size simulation to evaluate whether the energy storage layout plan meets the constraint conditions;

[0048] Judge whether the iteration termination condition is met. If so, output the optimal energy storage layout plan; if not, perform iteration through mutation and crossover operations to obtain the offspring plan, and use dynamic step-size simulation to evaluate whether the energy storage layout plan meets the constraint conditions; select the better plan between the parent plan and the offspring plan to enter the next generation; repeat the above iteration and selection process until the iteration termination condition is met.

[0049] As an alternative solution, the process of selecting the better plan between the parent plan and the offspring plan to enter the next generation is as follows:

[0050] If both the parent plan and the offspring plan meet the constraint conditions, select the plan with a smaller objective function;

[0051] If among the parent plan and the offspring plan, one plan meets all the constraint conditions and the other plan does not meet all the constraint conditions, select the plan that meets all the constraint conditions;

[0052] If both the parent plan and the offspring plan do not meet all the constraint conditions, select the plan with fewer violations of the constraint conditions.

[0053] In some other embodiments, the following technical solution is adopted:

[0054] An energy storage layout optimization system for a new energy power station cluster includes:

[0055] A data acquisition module for acquiring the historical output data of the new energy power station;

[0056] A massive scenario acquisition module for determining the structure of the first tree of the optimal vine Copula based on the mean shift clustering algorithm, and then determining the optimal vine Copula structure, determining the optimal Copula function based on the optimal vine Copula structure, sampling the optimal vine Copula function and performing inverse transformation to obtain massive scenarios of the output of multiple new energy power stations at different time sections;

[0057] A typical scenario acquisition module for reducing the massive scenarios of the output of the new energy power station at each time section to obtain typical scenarios of new energy output, and generating typical operation scenarios of the new power system by splicing in the time dimension;

[0058] A model construction module is used to construct energy storage economy evaluation indicators, peak shaving evaluation indicators, the ability indicators for suppressing the time-series volatility of new energy output, the ability indicators for suppressing the ramping of new energy, and frequency active support ability indicators respectively under each typical operation scenario; and to construct a mathematical model for the energy storage layout optimization problem considering grid-connected and network-forming energy storage based on the constructed indicators.

[0059] A model solving module is used to iteratively solve the mathematical model based on the dynamic step-size simulation and differential evolution algorithm to obtain the optimal energy storage layout scheme.

[0060] In some other embodiments, the following technical solutions are adopted:

[0061] A terminal device includes a processor and a memory. The processor is used to implement instructions; the memory is used to store multiple instructions, and the instructions are adapted to be loaded and executed by the processor to perform the above-mentioned energy storage layout optimization method for new energy power station clusters.

[0062] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0063] (1) When constructing the typical operation scenarios of the power system, the present invention realizes the generation of new energy output scenarios based on mean shift clustering and vine Copula structure, and optimizes the vine Copula structure through the mean shift clustering algorithm, which can more accurately describe the spatial correlation characteristics of the output of new energy power stations.

[0064] (2) The present invention simplifies the massive scenarios of the output of new energy power stations under each time section through the K-means clustering algorithm. During the clustering process, on the basis of the currently determined optimal initial center points, new initial center points are added, and through continuous iterative processes, the final clustering result is obtained; it can minimize the number of scenarios after reduction on the premise of ensuring the effectiveness of scenario reduction.

[0065] (3) Considering that there are significant differences in costs and functional characteristics between grid-connected energy storage and network-forming energy storage, the present invention comprehensively considers these two different types of energy storage in the energy storage layout optimization, and simultaneously takes into account the economy of the layout scheme and the improvement effect on the system safety and stability during the energy storage layout, so as to improve the system safety and stability at the lowest economic cost.

[0066] (4) The present invention comprehensively considers the influence of the energy storage layout scheme on peak shaving evaluation indicators, the ability indicators for suppressing the time-series volatility of new energy output, the ability indicators for suppressing the ramping of new energy, and frequency active support ability indicators. The obtained energy storage layout scheme can improve the safety and stability of the power system in multiple aspects.

[0067] (5) The present invention proposes a combined optimization method for the energy storage layout of a new energy power station cluster based on dynamic step simulation and differential evolution algorithm, which can improve the dynamic simulation speed and the optimization solution efficiency.

[0068] Other features and advantages of the present invention will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of this aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 It is a schematic diagram of the energy storage layout optimization method for a new energy power station cluster in an embodiment of the present invention;

[0070] Figure 2 It is a schematic diagram of the optimal vine Copula tree 1 structure in an embodiment of the present invention;

[0071] Figure 3 It is an example diagram of determining the optimal vine Copula structure based on the mean shift clustering algorithm in an embodiment of the present invention;

[0072] FIG. 4(a) and FIG. 4(b) are respectively schematic diagrams of the correlation between historical data and generated data;

[0073] Figure 5 It is a flow chart of the new energy output scenario reduction based on the improved K-means clustering algorithm in an embodiment of the present invention;

[0074] Figure 6 It is the daily load curve before and after the participation of energy storage in an embodiment of the present invention;

[0075] Figure 7 It is a schematic diagram of the composition of the output power of the energy storage power station in an embodiment of the present invention;

[0076] Figure 8 It is the system frequency response curve before and after the active support of the energy storage participation frequency in an embodiment of the present invention;

[0077] Figure 9 It is a schematic diagram of the process of model solution based on dynamic step simulation and differential evolution algorithm in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0078] It should be noted that the following detailed description is illustrative and is intended to provide further description of the present application. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present application belongs.

[0079] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they specify the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0080] Embodiment 1

[0081] In one or more embodiments, an energy storage layout optimization method for a new energy power station cluster is disclosed. First, a new energy output scenario generation method based on mean shift clustering and vine Copula fusion is proposed. The vine Copula structure is optimized by the mean shift clustering algorithm to more accurately describe the spatial correlation characteristics of the new energy power station output. Second, to improve the scenario reduction efficiency, a new energy output scenario reduction method based on an improved clustering algorithm is proposed. The initial clustering points of the K-means clustering algorithm are accurately determined to improve the scenario reduction effect and obtain typical new energy output scenarios. Then, when considering the energy storage layout optimization problem of the new energy power station cluster, evaluation indexes for energy storage economy, peak shaving, suppressing the time-series volatility of new energy output, suppressing new energy ramp-up, and frequency active support ability are constructed. On this basis, a mathematical model for the energy storage layout optimization problem is constructed. Finally, since the optimization problem involves a large number of time-domain simulations and the time-consuming of time-domain simulations results in low optimization solution efficiency, a combined optimization method for the energy storage layout of the new energy power station cluster based on dynamic step-size simulation and differential evolution algorithm is proposed to obtain the optimal energy storage layout scheme.

[0082] As a specific implementation scheme, in combination with Figure 1 , the method of this embodiment specifically includes the following processes:

[0083] S101: Obtain the historical output data of the new energy power station.

[0084] Assume that there are m new energy power stations in total, and obtain the historical output data of the m new energy power stations.

[0085] S102: Determine the structure of the first tree of the optimal vine Copula based on the mean shift clustering algorithm, and then determine the optimal vine Copula structure. Determine the optimal Copula function based on the optimal vine Copula structure, sample the optimal vine Copula function and perform inverse transformation to obtain a large number of scenarios of the output of multiple new energy power stations at different time sections.

[0086] Affected by factors such as landforms, there are intricate spatial correlation relationships among the outputs of new energy power stations. The vine Copula model is suitable for joint probability distribution modeling of multiple random variables, while the traditional C-vine or D-vine Copula structures are difficult to accurately describe the output correlation relationships among multiple new energy power stations. Therefore, based on the mean shift clustering algorithm, the optimal structure of the first tree of the vine Copula is determined, and then the optimal vine Copula structure is determined to generate a large number of new energy output scenarios.

[0087] Specifically, the specific process of determining the optimal vine Copula structure in this embodiment is as follows:

[0088] S1021: Assume that there are m new energy power stations in total, and calculate the Pearson correlation coefficient matrix P of the outputs among the power stations:

[0089]

[0090] In formula (1), P is an m-dimensional symmetric square matrix, and a ij is the Pearson correlation coefficient of the outputs between the i-th and j-th new energy power stations.

[0091] S1022: Define the spatial correlation coordinate xi of the i-th new energy power station i as the i-th row of the correlation coefficient matrix.

[0092] In this embodiment, the correlation coefficient matrix contains the correlation relationships among random variables. Therefore, the correlation coefficient is used to find the optimal vine Copula structure. Define the spatial correlation coordinate xi of the i-th new energy power station i as:

[0093] xi i = [a i1 , a i1 ,..., a im (2)

[0094] In formula (2), xi i is exactly the i-th row of the correlation coefficient matrix P. After representing the spatial correlation coordinates of each new energy power station, the new energy power stations can be clustered by the mean shift clustering algorithm.

[0095] S1023: Cluster the new energy power stations by the mean shift clustering algorithm based on the spatial correlation coordinates.

[0096] Specifically, the mean shift clustering algorithm can make each sample point quickly shift to the nearby dense area without the need to specify the number of clusters in advance. There are m spatial correlation coordinates xi i (i = 1,..., m) of the new energy power stations. For the spatial correlation coordinate x, its offset vector is:

[0097]

[0098] In formula (3), S x represents the high-dimensional spherical region around the spatial correlation coordinate x of the new energy power station, and s is the number of spatial correlation coordinates of the new energy power stations included in S x

[0099] Use the Gaussian kernel function to measure the weight of each x i i The Gaussian weight of x is:

[0100]

[0101] In formula (4), G(·) represents the Gaussian weight, h is the bandwidth, ||x i - x|| represents the Euclidean distance between x i and x. The smaller the bandwidth and the distance between nodes, the larger the Gaussian weight of x i

[0102] After introducing the Gaussian weight, the calculation formula for the x offset vector is:

[0103]

[0104] The density center point x' after one offset can be expressed as:

[0105]

[0106] Set d 1 and d 2 as the iteration stop threshold and the aggregation threshold respectively. During the drift process, each power station needs to determine the corresponding clustering center point through density center iteration. When the offset distance of each density center update is less than d 1 , the density center iteration process of this power station ends. During the aggregation process, when the distance between two clustering center points is less than d 2 , they are considered to belong to the same class and are aggregated. Thus, the clustering of new energy power stations is achieved.

[0107] S1024: Assume that the number of clustering categories after clustering is n. For each power station in each category, a D-vine structure is used for connection. Then, any two power stations are selected from different categories for vine structure connection to achieve the selection of the optimal vine Copula tree 1 structure, as Figure 2 shown.

[0108] The first tree has a great influence on the modeling effect of the vine Copula. After determining the structure of the first tree, the D-vine structure is used for the subsequent tree structures, so as to obtain the optimal vine Copula structure.

[0109] ​​​In this embodiment, the optimal Copula function is determined based on the optimal vine Copula structure. The specific process is as follows:

[0110] In the vine Copula structure, each edge connecting two nodes corresponds to a Copula function. There are five commonly used Copula functions: normal Copula, Clayton Copula, Gumbel Copula, t Copula, and Frank Copula. For the determination of the optimal Copula function of two variables, first, the marginal distributions of historical data are obtained through the empirical distribution function. The empirical distribution function is:

[0111]

[0112] In Equation (7), F i (x) is the empirical distribution function of the output of the i-th new energy power station, N h is the number of samples of historical data, I [条件] is the indicator function, which takes the value of 1 if the condition is satisfied, otherwise 0.

[0113] Secondly, the parameters of each Copula function are obtained based on the maximum likelihood estimation. Then, the Euclidean distance between the estimated Copula function and the empirical Copula function is calculated, and the Copula function with the minimum Euclidean distance is selected. The mathematical expression of the empirical Copula function is:

[0114]

[0115] Among them, N h is the number of samples of historical output data of the new energy power station, C e is the empirical Copula function, x 1,i is the output value of the i-th historical sample of the first new energy power station, x 2,i is the output value of the i-th historical sample of the second new energy power station, u i =F i (x) is the value of the empirical distribution function of the output of the i-th new energy power station, C n is the Copula function obtained by maximum likelihood estimation.

[0116] The Euclidean distance between the empirical Copula and the Copula function is defined as:

[0117]

[0118] The Euclidean distance between the empirical Copula and the Copula function reflects the goodness or badness of the fitting effect of the Copula function on the sample data distribution. The smaller the Euclidean distance between the two, the better the fitting effect of the Copula function on the historical data distribution. Therefore, the Copula function corresponding to the smallest Euclidean distance is selected as the optimal Copula function for two variables.

[0119] Under the obtained optimal vine Copula structure, the optimal Copula function corresponding to each edge of the vine Copula is determined in turn, so as to obtain the optimal vine Copula function. Sampling the optimal vine Copula and performing inverse transformation to generate a large number of scenarios of the output of multiple new energy power stations at the same time section.

[0120] Establish a vine Copula model for each time section according to the above method. After sampling and inverse transformation, the output scenarios of the new energy power stations under multiple time sections can be obtained.

[0121] This embodiment realizes the generation of new energy output scenarios based on mean shift clustering and vine Copula structure. By optimizing the vine Copula structure through the mean shift clustering algorithm, the spatial correlation characteristics of the output of new energy power stations can be described more accurately. Taking 5 wind farms in a certain place as an example, the effectiveness of the proposed new energy output scenario generation method in this embodiment is verified.

[0122] The data is the output data of 5 wind farms at the same time section every day for 427 days.

[0123] Determine the optimal vine Copula structure based on the mean shift clustering algorithm as Figure 3 shown. Generate 20,000 groups of wind farm output operation scenarios based on the optimal vine Copula structure. The correlations between the historical data and the generated data are shown in Figures 4(a) and 4(b) respectively. It can be seen that in the historical data, when the correlation between two variables is strong, the correlation of the generated data is also strong; when the correlation between two variables in the historical data is weak, the correlation of the generated data is also weak. Therefore, the optimal vine Copula structure can fully consider the spatial correlation between the outputs of each power station.

[0124] Table 1 gives the comparison of the marginal distribution characteristics of the historical data and the generated data of the output of each wind farm.

[0125] Table 1 Comparison of the marginal distribution characteristics of the historical data and the generated data of the output of each wind farm

[0126]

[0127] As can be seen from Table 1, the generated data of each wind farm is very close to the historical data in terms of both the mean value and the standard deviation, indicating that the generated data can meet the spatial correlation of the output of multiple wind farms while conforming to their respective marginal distribution characteristics, verifying the effectiveness of the proposed method for generating output scenarios of multiple new energy power stations.

[0128] S103: Simplify the massive scenarios of the output of new energy power stations at each time section to obtain typical scenarios of new energy output, and then generate typical operation scenarios of the new power system through splicing in the time dimension.

[0129] If the generated massive scenarios are directly used in the subsequent combined optimization of energy storage layout, the excessive number of scenarios will lead to a large computational burden and a long optimization time. Therefore, it is necessary to obtain typical output scenarios from the massive new energy output scenarios through scenario reduction. The K-means clustering algorithm is suitable for scenario reduction of large-scale power systems. However, the clustering effect of this clustering algorithm is greatly affected by the selection of the initial clustering points. Therefore, in this embodiment, the initial points of the clustering algorithm are optimized.

[0130] The basic idea of the K-means clustering algorithm is to find a partitioning scheme of k clusters to minimize the loss function corresponding to the clustering result. The loss function is defined as the sum of the squared errors of each sample point to the center point of the cluster to which it belongs:

[0131]

[0132] In Equation (10), s i is the i-th output scenario of multiple new energy power stations generated by the vine Copula model at the same time section, c i is the clustering category to which the sample s i belongs, u ci represents the clustering center point corresponding to this clustering category, and N g is the number of output scenarios of new energy power stations generated by the vine Copula at the same time section.

[0133] The selection of the initial center point has a great impact on the result of the clustering algorithm. In this embodiment, a greedy algorithm is used to select appropriate initial center points, which can minimize the number of scenarios after reduction as much as possible while ensuring the effectiveness of scenario reduction.

[0134] Combined with Figure 5 , the specific process is as follows:

[0135] S1031: Initialize N r ;

[0136] S1032: Traverse any N r scenarios as the initial center points, and take the N rOne scenario is used as the initial clustering center point of the K-means clustering algorithm; randomly select one scenario from the remaining scenarios and N r initial clustering center points as the new initial clustering center points for clustering; select the N r +1 samples with the best clustering results as the initial points of the K-means clustering algorithm;

[0137] S1033: Determine whether N r reaches the number of typical scenarios of new energy output. If it reaches, output the clustering result. Otherwise, let N r = N r +1, repeat the process of S1032 until N r reaches the number N rset of typical scenarios of new energy output, and output the clustering result as the typical scenarios of new energy output under each time section.

[0138] For the new energy output scenarios generated in multiple time sections, the above scenario reduction method is adopted to obtain the typical scenarios of new energy output in multiple time sections. The scenarios in multiple time sections are spliced in the time dimension to obtain the typical time series scenarios of new energy output of multiple new energy power stations, and then the typical operating scenarios of the new power system are generated.

[0139] S104: Under each typical operating scenario, construct the energy storage economy evaluation index, peak regulation evaluation index, ability index to suppress the time series volatility of new energy output, ability index to suppress the new energy ramp-up ability, and frequency active support ability index respectively; based on the constructed indexes, construct a mathematical model for the energy storage layout optimization problem considering the grid-connected type and grid-forming type.

[0140] In this embodiment, the specific index construction method is as follows:

[0141] (1) Energy storage economy evaluation index C EES :

[0142]

[0143] In formula (11), N EES is the number of energy storage power stations, c 1 is the unit capacity cost (yuan / kW·h) of the grid-connected energy storage, and c 2 is the unit power cost (yuan / kW·h) of the grid-forming energy storage; E i E1 and E i E2 are the capacities of the grid-connected and grid-forming energy storages configured for the i-th new energy power station respectively.

[0144] (2) Energy storage participation in system peak regulation index Y peek :

[0145] The double uncertainties of power sources and loads exacerbate the peak-valley difference of the power grid. By allowing energy storage to participate in system peak shaving, the energy storage discharges during high daily loads and charges during low daily loads, which can effectively reduce the peak shaving pressure of the system. As Figure 6 shown.

[0146] Figure 6 In 1 , the blue curve L 2 (t) represents the load power curve of the system when the energy storage does not participate in system peak shaving, and the red curve L 1 (t) represents the load power curve of the system after the energy storage participates in peak shaving. P 2 represents the peak shaving boundary, and P Figure 6 represents the valley filling boundary. The area of the shaded part in

[0147] can represent the magnitude of the peak shaving capacity of the energy storage. The larger the area of the shaded part, the stronger the ability of the energy storage to participate in system peak shaving. peek Therefore, the peak shaving index Y

[0148]

[0149] In Equation (12), I [条件] is an indicator function. If the condition is satisfied, it takes the value of 1; otherwise, it takes the value of 0. t is the time length of the time series scenario, Y peek corresponds to the area of the shaded part in the above figure, and the larger the value of Y peek , the larger the corresponding area of the shaded part, and the stronger the ability of the corresponding energy storage to participate in system peak shaving.

[0150] (3) Energy storage smoothing capacity evaluation index Y 1 for the time series volatility of new energy output:

[0151] For each energy storage power station, the output power composition is as Figure 7 shown. Figure 7 In wind , P pv is the output power of the wind farm, P i total is the total power of the corresponding tie line of the i-th energy storage power station, satisfying power balance, as shown in Equation (13):

[0152]

[0153] The energy storage smoothing capacity evaluation index can be defined as the tie line power fluctuation. The smaller the fluctuation, the smaller the index, and the better the stability of the system, as shown in Equation (14):

[0154]

[0155] In Equation (14), Y1 is the evaluation index for the energy storage smoothing ability, t is the length of the time series scenario, and P i total (j) is the total power of the tie line at the j-th moment of the i-th energy storage station.

[0156] (4) Quantification index for energy storage to smooth new energy ramp:

[0157] The output of new energy has large fluctuations. A large change in the output of new energy in a short period is called a new energy ramp event. The ramp event is very harmful to the power grid. Energy storage can effectively smooth the new energy ramp by quickly charging and discharging in response to the rapid power fluctuations of new energy.

[0158] In this embodiment, the ramp rate of the tie line power of the new energy station is used to quantify the ability of energy storage to smooth the new energy ramp. The ramp rate is defined as:

[0159]

[0160] In formula (15), Y i ramp (x) is the output ramp rate of the i-th new energy station at the x-th moment, Δt is the time interval, and P i total (x) is the total power of the tie line of the i-th new energy station at the x-th moment. In particular, when Δt takes the value of 1, Y i ramp (x) reflects the power ramp rate of the new energy station at adjacent moments. Formula (15) can be written as:

[0161] Y i ramp (x) = P i total (x + 1) - P i total (x) (16)

[0162] Compared with the situation where energy storage does not participate in smoothing the new energy ramp, when energy storage participates in smoothing the new energy ramp, the absolute value of Y i ramp (x) will become smaller, that is, the new energy ramp rate decreases. The smaller the absolute value of Y i ramp (x), the better the effect of energy storage in smoothing the new energy ramp.

[0163] (5) Evaluation index for the active frequency support ability of energy storage:

[0164] After an active power deficit occurs in the power system, the system frequency drops, which may cause system frequency stability problems and then trigger other system stability problems. Energy storage stations can improve the frequency safety and stability of the system after an active power deficit by participating in the active support of the system frequency. After a large active power deficit occurs in the system, the schematic diagram of the system frequency response when the energy storage power station participates in the active support of the frequency is as shown in Figure 8 shown; Figure 8 In 1 , t 2 is the moment when the active power deficit occurs in the system, t 1 is the end moment of the primary frequency regulation of the system. The black curve f 2 (t) represents the frequency response curve of the system after the energy storage station participates in the active support of the frequency, and the red curve f fsp (t) represents the frequency response curve of the system when the energy storage station does not participate in the active support of the frequency. Therefore, the evaluation index Y

[0165]

[0166] In this embodiment, if E i E1 , E i E2 (i = 1, 2,..., m) are the capacities of the grid-connected and network-forming energy storage configured by the i-th new energy station respectively, then the mathematical model of the energy storage layout combination optimization problem can be expressed as:

[0167]

[0168] In formula (18), c 1 is the cost per unit capacity of the grid-connected energy storage (yuan / kW·h), c 2 is the cost per unit power of the network-forming energy storage (yuan / kW·h); E i E1 , E i E2 are the capacities of the grid-connected and network-forming energy storage configured by the i-th new energy station respectively; m is the number of new energy stations, Y peek_set is the lower limit of the defined energy storage peak shaving index, P i wind (t) is the active power of the wind power of the i-th new energy station at time t, P i pv (t) is the active power of the photovoltaic power of the i-th new energy station at time t, p i E1 (t) is the active power of the grid-connected energy storage of the i-th new energy station at time t, p i E2(t) is the active power of the i-th new energy power station's grid-forming energy storage at time t, P i base is the reported reference capacity, δ is the maximum allowable boundary of the tie-line power fluctuation, δ > 0, Y i ramp_l 、Y i ramp_h are the minimum and maximum new energy ramp rates of the i-th new energy power station, t 1 is the moment when the active power deficit occurs, t 2 is the moment when the system's primary frequency regulation ends and returns to the steady state, d f is the minimum frequency active support capacity boundary of the energy storage, P i E1 is the power upper bound of the i-th new energy power station's grid-following energy storage, P i E2 is the power upper bound of the i-th new energy power station's grid-forming energy storage, SOC i 1 (t) is the state of charge of the i-th new energy power station's grid-following energy storage at time t, SOC i 1m (t) is the minimum state of charge of the i-th new energy power station's grid-following energy storage, SOC i 1M (t) is the maximum state of charge of the i-th new energy power station's grid-following energy storage, SOC i 2 (t) is the state of charge of the i-th new energy power station's grid-forming energy storage at time t, SOC i 2m (t) is the minimum state of charge of the i-th new energy power station's grid-forming energy storage, SOC i 2M (t) is the maximum state of charge of the i-th new energy power station's grid-forming energy storage.

[0169] S105: Based on the dynamic step-size simulation and differential evolution algorithm, iteratively solve the mathematical model to obtain the optimal energy storage layout scheme.

[0170] From the mathematical model of the energy storage layout combination optimization problem, it can be seen that there are many constraint conditions in this optimization problem, and there are also non-linear constraints. It is necessary to judge whether the constraint conditions are satisfied through time-domain simulation. However, time-domain simulation requires solving a large number of differential-algebraic simulation groups, and a large number of time-domain simulations are very time-consuming, which reduces the optimization efficiency. In this embodiment, a dynamic step-size simulation strategy is proposed to improve the dynamic simulation speed and the optimization solution efficiency. The execution method of the dynamic simulation step-size strategy is as follows:

[0171] First, initialize the simulation stage and the fault-free simulation stage. Since the states of the system are stable, a large step size is used for time-domain simulation.

[0172] When disturbance occurs or control is applied, the system is in a non-steady-state state, and dynamic simulation is performed with a small step size to more accurately characterize the dynamic behavior of the system.

[0173] Finally, when the system transitions to a stable state again, a large step size is used again for dynamic simulation.

[0174] The flow chart of the rapid optimization method for energy storage layout based on dynamic simulation step size and differential evolution algorithm is as follows: Figure 9 As shown, the specific process is as follows:

[0175] S1052: Initialize the energy storage layout plan, and use dynamic step size simulation to evaluate whether the energy storage layout plan meets the constraint conditions;

[0176] S1053: Determine whether the iteration termination condition is met. If so, output the optimal energy storage layout plan; if not, iterate through mutation and crossover operations to obtain the child plan, and use dynamic step simulation to evaluate whether the energy storage layout plan meets the constraint conditions; select the better plan between the parent plan and the child plan to enter the next generation; repeat the above iteration and selection process until the iteration termination condition is met.

[0177] In this embodiment, the termination condition can be set to whether the current number of evolutions reaches the set number of evolutions. In the selection operation, the better solution between the parent solution and the offspring solution is selected to enter the next generation. Specifically, the comparison method of the advantages and disadvantages of the parent solution and the offspring solution is:

[0178] (1) If both the parent solution and the child solution satisfy the constraints, the solution with the smaller objective function is selected;

[0179] (2) If the constraints of one of the parent and child solutions are all met, but the constraints of the other solution cannot be fully met, then the solution that satisfies all constraints is selected;

[0180] (3) If the constraints of both the parent and child solutions cannot be fully satisfied, the solution with fewer violated constraints is selected.

[0181] The combined optimization method of energy storage layout of new energy station cluster based on dynamic step simulation and differential evolution algorithm can improve the dynamic simulation speed and the optimization solution efficiency.

[0182] Embodiment 2

[0183] In one or more embodiments, a system for optimizing energy storage layout for a new energy station cluster is disclosed, including:

[0184] Data acquisition module, used to obtain historical output data of new energy stations;

[0185] A massive scenario acquisition module, which is used to determine the structure of the first tree of the optimal vine Copula based on the mean shift clustering algorithm, and then determine the optimal vine Copula structure. Based on the optimal vine Copula structure, it determines the optimal Copula function, samples the optimal vine Copula function and performs inverse transformation to obtain massive scenarios of the output of multiple new energy power stations at different time sections.

[0186] A typical scenario acquisition module, which is used to simplify the massive scenarios of the output of new energy power stations at each time section to obtain typical scenarios of new energy output, and generate typical operation scenarios of the new power system through splicing in the time dimension.

[0187] A model construction module, which is used to construct an energy storage economy evaluation index, a peak shaving evaluation index, an ability index for suppressing the time series volatility of new energy output, an evaluation index for suppressing the ramp-up ability of new energy, and a frequency active support ability index under each typical operation scenario respectively; based on the constructed indexes, it constructs a mathematical model for the energy storage layout optimization problem considering grid-connected and network-forming types.

[0188] A model solving module, which is used to iteratively solve the mathematical model based on the dynamic step-size simulation and differential evolution algorithm to obtain the optimal energy storage layout scheme.

[0189] It should be noted that the specific implementation methods of the above modules have been described in detail in Embodiment 1, and will not be elaborated here.

[0190] Embodiment 3

[0191] In one or more embodiments, a terminal device is disclosed, which includes a processor and a memory. The processor is used to implement instructions; the memory is used to store multiple instructions, and the instructions are suitable for being loaded and executed by the processor to perform the energy storage layout optimization method for new energy power station clusters described in Embodiment 1.

[0192] It should be understood that in this embodiment, the processor may be a central processing unit CPU, and the processor may also be other general-purpose processors, digital signal processors DSP, application-specific integrated circuits ASIC, off-the-shelf programmable gate arrays FPGA or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0193] The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A part of the memory may also include a non-volatile random memory. For example, the memory may also store information about the device type.

[0194] In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in the processor or the instructions in the form of software.

[0195] Although the specific implementation manners of the present invention have been described above in conjunction with the accompanying drawings, it is not a limitation on the protection scope of the present invention. Those skilled in the art should understand that, based on the technical solutions of the present invention, various modifications or deformations that can be made by those skilled in the art without creative efforts are still within the protection scope of the present invention.

Claims

1. A method for optimizing energy storage layout for a new energy station cluster, characterized in that: include: Obtain historical output data of new energy stations; Based on the mean shift clustering algorithm, the first tree structure of the optimal vine copula is determined, and then the optimal vine copula structure is determined. Based on the optimal vine copula structure, the optimal copula function is determined. The optimal vine copula function is sampled and inversely transformed to obtain a large number of scenarios of the output of multiple renewable energy stations under different time sections. Simplify the massive scenarios of renewable energy station output at each time section to obtain typical scenarios of renewable energy output, and then generate typical operation scenarios of new power systems through splicing in the time dimension; In each typical operation scenario, we construct energy storage economic evaluation indicators, peak load evaluation indicators, ability indicators to smooth out the time series fluctuation of new energy output, ability indicators to smooth out the ramping of new energy, and active frequency support ability evaluation indicators. Based on the constructed indicators, we construct a mathematical model for energy storage layout optimization problems that consider both grid-following and grid-building types. Based on dynamic step-size simulation and differential evolution algorithm, the mathematical model is iteratively solved to obtain the optimal energy storage layout solution.

2. The energy storage layout optimization method for a new energy station cluster according to claim 1, characterized in that: Determine the optimal vine copula structure. The specific process is as follows: Assuming there are m renewable energy stations, calculate the Pearson correlation coefficient matrix of the output between the stations; Define the spatial correlation coordinate x of the i-th new energy station i is the i-th row of the correlation coefficient matrix; Based on spatial correlation coordinates, the new energy stations are clustered using the mean shift clustering algorithm; Assume that the number of clustering categories after clustering is n. Each station in each category is connected by D-vine structure. Then, any two stations in different categories are selected to connect with the vine structure to achieve the selection of the optimal vine Copula tree 1 structure. The D-vine structure is used for the subsequent tree structures to obtain the optimal vine Copula structure.

3. The energy storage layout optimization method for a new energy station cluster according to claim 1, characterized in that: The optimal Copula function is determined based on the optimal vine Copula structure. The specific process is as follows: Determine an empirical distribution function of the output of each new energy station, and determine an empirical Copula function based on the empirical distribution function; Each edge in the optimal vine Copula structure corresponds to a Copula function, and the parameters of each Copula function are obtained based on maximum likelihood estimation; Calculate the Euclidean distance between each Copula function corresponding to the current edge and the empirical Copula function, and select the Copula function with the smallest Euclidean distance as the optimal Copula function for the edge; The optimal Copula function corresponding to each edge of the optimal vine Copula structure is determined in turn, thereby obtaining the optimal vine Copula function.

4. The energy storage layout optimization method for a new energy station cluster according to claim 1, characterized in that: The massive scenarios of new energy station output at each time section are simplified. The specific process is as follows: (1) Initialize N r ; (2) Traverse any N r scenes as the initial center point, and the N scenes with the best clustering results are r scenes as the initial clustering center points of the K-means clustering algorithm; randomly select one scene and N from the remaining scenes r The initial cluster center points are used as new initial cluster center points for clustering; Select the N with the best clustering result r +1 sample as the initial point of the K-means clustering algorithm; (3) Determine N r Whether the number of typical scenarios of new energy output is reached, if so, output the clustering results, otherwise, let N r =N r +1, repeat the process of (2) until N r The number of typical scenarios of new energy output is reached, and the clustering results are output as the typical scenarios of new energy output in each time section.

5. The energy storage layout optimization method for a new energy station cluster according to claim 1, characterized in that: Constructing energy storage economic evaluation index C EES , specifically: Among them, N EES is the number of energy storage power stations, c1 is the unit capacity cost of grid-connected energy storage, and c2 is the unit power cost of grid-connected energy storage; E i E1 、E i E2 The capacity of grid-following and grid-building energy storage is configured for the i-th new energy station respectively; Or, construct the peak load evaluation index Y peek , specifically: Among them, I [条件] is an indicative function. If the condition is met, the value is 1, otherwise the value is 0. t is the time length of the timing scenario. L1(t) represents the net load power curve of the system when the energy storage does not participate in the peak regulation of the system. L2(t) represents the net load power curve of the system after the energy storage participates in the peak regulation. P1 represents the peak shaving boundary, and P2 represents the valley filling boundary. Alternatively, construct the capability index Y1 for smoothing the time series fluctuation of new energy output, specifically: Among them, P i total (j) is the total power of the interconnection line at the i-th energy storage station at the j-th moment, P base,i is the reported baseline capacity; Alternatively, construct an evaluation index for the ability to stabilize the ramp-up of new energy, specifically: Among them, Y i ramp (x) is the output ramp rate of the i-th renewable energy station at time x, Δt is the time interval, P i total (x) is the total power of the interconnection line at the i-th new energy station at time x; Alternatively, construct frequency active support capability evaluation indicators, specifically: Among them, t1 is the time when the system active power shortage occurs, t2 is the time when the system primary frequency regulation ends, f1(t) represents the frequency response curve of the system after the energy storage station participates in the active frequency support, and f2(t) represents the frequency response curve of the system when the energy storage station does not participate in the active frequency support.

6. The energy storage layout optimization method for a new energy station cluster as claimed in claim 5, characterized in that: Based on the constructed indicators, a mathematical model for the optimization of energy storage layout considering the grid-following type and grid-building type is constructed, specifically: Among them, c1 is the unit capacity cost of grid-following energy storage, c2 is the unit power cost of grid-building energy storage; E i E1 、E i E2 are the capacity of grid-following and grid-building energy storage configured for the i-th new energy station respectively; m is the number of new energy stations, Y peek_set is the lower limit of energy storage peak load regulation index, P i wind (t) is the active power of wind power of the i-th renewable energy station at time t, P i pv (t) is the active power of the photovoltaic power station of the i-th new energy station at time t, p i E1 (t) is the active power of the i-th new energy station and grid-type energy storage at time t, p i E2 (t) is the active power of the grid-connected energy storage of the i-th new energy station at time t, P base,i is the reported benchmark capacity, δ is the maximum allowable limit of the interconnection line power fluctuation, δ>0, Y i ramp_l , Y i ramp_h are the minimum and maximum values ​​of the new energy ramp rate of the i-th new energy station, d f is the boundary of the active support capability with the maximum energy storage and minimum frequency; P i E1 is the upper bound of the power of the i-th new energy station and grid-type energy storage, P i E2 The upper bound of the grid-type energy storage power of the i-th new energy station; SOC i 1(t) is the charge state of the i-th new energy station and grid-type energy storage at time t, SOC i 1m (t) is the minimum state of charge of the i-th new energy station and grid-type energy storage, SOC i 1M (t) is the maximum state of charge of the i-th new energy station and grid-type energy storage, SOC i 2(t) is the state of charge of the grid-connected energy storage of the i-th new energy station at time t, SOC i 2m (t) is the minimum state of charge of the grid-connected energy storage of the i-th new energy station, SOC i 2M (t) is the maximum charge state of the grid-type energy storage of the i-th new energy station.

7. The energy storage layout optimization method for a new energy station cluster according to claim 1, characterized in that: Based on dynamic step-size simulation and differential evolution algorithm, the mathematical model is iteratively solved to obtain the optimal energy storage layout solution, which is as follows: Initialize the energy storage layout plan and use dynamic step size simulation to evaluate whether the energy storage layout plan meets the constraints; Determine whether the iteration termination condition is met, and if so, output the optimal energy storage layout plan; If not, the offspring solution is obtained through iteration through mutation and crossover operations, and dynamic step size simulation is used to evaluate whether the energy storage layout solution meets the constraints; The better solution between the parent solution and the offspring solution is selected to enter the next generation; the above iteration and selection process is repeated until the iteration termination condition is met.

8. The energy storage layout optimization method for a new energy station cluster as claimed in claim 7, characterized in that: The better solution between the parent solution and the offspring solution is selected to enter the next generation. The specific process is as follows: If both the parent solution and the child solution satisfy the constraints, the solution with the smaller objective function is selected; If the constraints of one of the parent and child solutions are all met, but the constraints of the other solution cannot be fully met, then the solution that satisfies all constraints is selected; If the constraints of both the parent and child solutions cannot be fully satisfied, the solution with fewer violated constraints will be selected.

9. An energy storage layout optimization system for new energy station clusters, characterized in that: include: Data acquisition module, used to obtain historical output data of new energy stations; Massive scene acquisition module, used to determine the first tree structure of the optimal vine Copula based on the mean shift clustering algorithm, and then determine the optimal vine Copula structure, determine the optimal Copula function based on the optimal vine Copula structure, sample the optimal vine Copula function and perform inverse transformation, and obtain massive scenes of multiple renewable energy station outputs under different time sections; The typical scenario acquisition module is used to simplify the massive scenarios of new energy station output at each time section, obtain the typical scenarios of new energy output, and splice them through the time dimension to generate the typical operation scenarios of the new power system; The model building module is used to construct energy storage economic evaluation indicators, peak load evaluation indicators, ability indicators for smoothing the timing fluctuation of new energy output, evaluation indicators for smoothing the ramping ability of new energy, and evaluation indicators for active frequency support ability in each typical operation scenario; based on the constructed indicators, a mathematical model for energy storage layout optimization problems considering both grid-following and grid-building types is constructed; The model solving module is used to iteratively solve the mathematical model based on dynamic step simulation and differential evolution algorithm to obtain the optimal energy storage layout solution.

10. A terminal device, comprising a processor and a memory, wherein the processor is used to implement instructions; and the memory is used to store multiple instructions, characterized in that: The instructions are suitable for being loaded by a processor and executing the energy storage layout optimization method for a new energy station cluster as described in any one of claims 1-8.

Images