Large-scale energy storage power station evaluation method and system based on multi-dimensional data fusion

Abstract

The invention provides a large-scale energy storage power station evaluation method and system based on multi-dimensional data fusion, and is applied to the technical field of energy management. The method comprises the following steps: calculating values of evaluation indexes corresponding to a plurality of evaluation angles based on multi-dimensional operation data of the large-scale energy storage power station, and fusing the values of the evaluation indexes corresponding to the plurality of evaluation angles into an original matrix of the evaluation indexes of the large-scale energy storage power station based on the energy storage power station; performing dimension reduction processing on the original matrix by using a principal component analysis method to obtain a dimension reduction matrix; performing association processing on the expert weight of each evaluation index in the dimension reduction matrix by using a grey relational degree method to obtain an index weight; and performing clustering processing on the value of each evaluation index in the dimension reduction matrix, and performing evaluation on each energy storage power station in each clustering cluster based on the value of each evaluation index in each clustering cluster in a clustering result and a corresponding index weight. According to the method, the problems that the evaluation process of the large-scale energy storage power station lacks an overall evaluation mode and the evaluation result is low in accuracy are solved.

Description

Technical Field

[0001] The present invention relates to the field of energy management technology, and in particular to a large-scale energy storage power station evaluation method and system based on multi-dimensional data fusion. Background Art

[0002] Energy storage technology involves converting energy in a form that is difficult to store into a more convenient or economically storable form, mainly referring to the storage of electrical energy. The stored energy can be used as emergency energy or for energy storage when the grid load is low. Energy storage technology is expected to become an important force in energy transformation. Energy storage can provide a variety of services such as peak regulation, frequency regulation, black start, and demand response support for grid operation. Its rapid response characteristics have greatly improved the flexibility, economy, and safety of traditional power systems. Based on this, energy storage power stations are produced. Energy storage power stations are equipment systems that store, convert, and release cyclic electrical energy through electrochemical batteries or electromagnetic energy storage media. Accurate and real-time evaluation of the operating status and adjustable capacity of energy storage power stations is an important prerequisite for effectively exerting the role of energy storage and ensuring the safe and reliable operation of energy storage power stations.

[0003] The data collected from energy storage power stations are characterized by large amounts of data and numerous attributes. Especially for large-scale energy storage power stations, there are many evaluation indicators, and the amount of data for each energy storage power station is very large, making it difficult to conduct a comprehensive evaluation. Most of the current energy storage power station evaluation methods use the Analytic Hierarchy Process (AHP) or the entropy weight method to determine the indicator weights. However, the AHP has problems such as subjectivity and ambiguity in the evaluation process, and the entropy weight method cannot handle situations where there is a strong correlation between indicators, and is prone to unreasonable weight distribution problems. Therefore, the current evaluation process of large-scale energy storage power stations lacks a comprehensive evaluation method, and the evaluation results are inaccurate. Summary of the invention

[0004] In order to overcome the problems of lack of overall evaluation method and low accuracy of evaluation results in the above-mentioned large-scale energy storage power station evaluation process, the present invention provides a large-scale energy storage power station evaluation method and system based on multi-dimensional data fusion.

[0005] On the one hand, the present invention provides a large-scale energy storage power station evaluation method based on multi-dimensional data fusion, comprising:

[0006] Calculate the values ​​of each evaluation index corresponding to multiple evaluation angles based on the acquired multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station, and fuse the values ​​of each evaluation index corresponding to the multiple evaluation angles into an original matrix of the evaluation index of the large-scale energy storage power station based on the energy storage power station;

[0007] Performing dimensionality reduction processing on the original matrix using principal component analysis to obtain a corresponding dimensionality reduction matrix;

[0008] The expert weights of the evaluation indicators in the dimension reduction matrix are correlated by using the grey correlation method to obtain the indicator weights of the evaluation indicators;

[0009] The values ​​of the evaluation indicators in the dimension reduction matrix are clustered, and the energy storage power stations in each cluster are evaluated based on the values ​​of the evaluation indicators in each cluster in the clustering result and the corresponding indicator weights.

[0010] Optionally, before calculating the values ​​of each evaluation index corresponding to multiple evaluation angles based on the acquired multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station, the method further includes:

[0011] Collecting the energy dimension operation data of each energy storage power station in the large-scale energy storage power station through the energy management system of the large-scale energy storage power station;

[0012] Collecting the operation data of the power regulation dimension of each energy storage power station in the large-scale energy storage power station through the energy storage converter of the large-scale energy storage power station;

[0013] The battery management system of the large-scale energy storage power station is used to collect the operating data of the battery dimension of each energy storage power station in the large-scale energy storage power station.

[0014] Optionally, the evaluation angles include energy efficiency evaluation angle, regulation performance evaluation angle, grid connection performance evaluation angle and reliability evaluation index;

[0015] The evaluation indicators corresponding to the grid-connected performance evaluation angle include an inertia response evaluation indicator, a primary frequency regulation evaluation indicator, an automatic generation control evaluation indicator, and an automatic voltage control evaluation indicator;

[0016] The evaluation indicators corresponding to the reliability evaluation angle include battery reliability evaluation indicators and power station reliability evaluation indicators; the battery reliability evaluation indicators include battery consistency indicators, battery health status and battery failure indicators, and the power station reliability evaluation indicators include power station outage coefficient, power station availability coefficient and power station utilization coefficient.

[0017] Optionally, the using principal component analysis method to perform dimensionality reduction processing on the original matrix to obtain a corresponding dimensionality reduction matrix includes:

[0018] Performing normalization processing on the original matrix to obtain a normalized matrix;

[0019] Obtaining a corresponding correlation coefficient matrix based on the correlation coefficients between the elements in the standardized matrix, and selecting a specified number of target eigenvalues ​​based on the eigenvalues ​​of the correlation coefficient matrix;

[0020] Determine a corresponding target eigenvector based on the target eigenvalue to obtain an eigenvector matrix;

[0021] The dimension reduction matrix is ​​obtained based on the normalized matrix and the eigenvector matrix.

[0022] Optionally, the normalizing the original matrix to obtain a normalized matrix includes:

[0023] Standardize each element of the original matrix based on the mean and standard deviation of each evaluation index in the original matrix to obtain a standardized matrix;

[0024] The step of obtaining the dimension reduction matrix based on the normalized matrix and the eigenvector matrix includes:

[0025] For each column of data in the standardized matrix, each element in each column of data is normalized based on the index mean of each column of data to obtain a sample matrix;

[0026] The product of the sample matrix and the eigenvector matrix is ​​used as the dimension reduction matrix.

[0027] Optionally, the grey correlation method is used to correlate the initial expert weights of the evaluation indicators in the dimension reduction matrix to obtain the indicator weights of the evaluation indicators, including:

[0028] Based on the weight score of each expert for each evaluation indicator in the dimensionality reduction matrix, an initial expert weight matrix corresponding to the dimensionality reduction matrix is ​​obtained, and each column of the initial expert weight matrix is ​​used as a comparison data column;

[0029] Screening out the maximum value of each column from the initial expert weight matrix to form a reference data column;

[0030] Based on the Mahalanobis distance between each comparison data column and the reference data column, an indicator weight of the evaluation indicator corresponding to each comparison data column is determined.

[0031] Optionally, the calculation formula of the indicator weight of the evaluation indicator is:

[0032]

[0033] Among them, ω k represents the index weight of the kth evaluation index, ω′ k Represents ω k The data before normalization, E0 represents the reference data column, E k represents the kth comparison data column, k=1, 2, ..., p, and p is the number of evaluation indicators in the dimension reduction matrix; D s represents the Mahalanobis distance.

[0034] Optionally, clustering the values ​​of the evaluation indicators in the dimension reduction matrix includes:

[0035] Performing maximum-minimum normalization processing on each element of the dimension reduction matrix to obtain a standardized dimension reduction matrix;

[0036] The row vectors of the standardized dimensionality reduction matrix are used as data points, and the distance between two row vectors is used as the distance between data points. The row vectors in the standardized dimensionality reduction matrix are clustered using a density-based clustering algorithm to obtain the center point of each cluster cluster and its coordinates; the center point of the cluster cluster is used to indicate the average score of each energy storage power station in the cluster cluster, and the coordinate of the center point in each evaluation index dimension is used to indicate the average score of all energy storage power stations in the corresponding cluster cluster for each evaluation index.

[0037] Optionally, the evaluating of each energy storage power station in each cluster based on the value of each evaluation index in each cluster in the clustering result and the corresponding index weight includes:

[0038] For each cluster, the product of the coordinates of the center point of the cluster in each evaluation index dimension and the corresponding index weight is accumulated to obtain a comprehensive score for each cluster;

[0039] The energy storage power stations within each cluster are evaluated based on the comprehensive score of each cluster.

[0040] On the other hand, the present invention also provides a large-scale energy storage power station evaluation system based on multi-dimensional data fusion, the system comprising:

[0041] A data fusion module is used to calculate the values ​​of each evaluation index corresponding to multiple evaluation angles based on the multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station obtained, and fuse the values ​​of each evaluation index corresponding to the multiple evaluation angles into an original matrix of the evaluation index of the large-scale energy storage power station based on the energy storage power station;

[0042] A dimension reduction module is used to perform dimension reduction processing on the original matrix using a principal component analysis method to obtain a corresponding dimension reduction matrix;

[0043] A weight determination module is used to perform correlation processing on the expert weights of each evaluation index in the dimension reduction matrix using a grey correlation method to obtain an index weight of each evaluation index;

[0044] The evaluation module is used to cluster the values ​​of the evaluation indicators in the dimension reduction matrix, and evaluate the energy storage power stations in each cluster based on the values ​​of the evaluation indicators in each cluster and the corresponding indicator weights in the cluster results.

[0045] Optionally, the system further comprises a data acquisition module, and the data acquisition module is used to:

[0046] Collecting the energy dimension operation data of each energy storage power station in the large-scale energy storage power station through the energy management system of the large-scale energy storage power station;

[0047] Collecting the operation data of the power regulation dimension of each energy storage power station in the large-scale energy storage power station through the energy storage converter of the large-scale energy storage power station;

[0048] The battery management system of the large-scale energy storage power station is used to collect the operating data of the battery dimension of each energy storage power station in the large-scale energy storage power station.

[0049] Optionally, the evaluation angles include energy efficiency evaluation angle, regulation performance evaluation angle, grid connection performance evaluation angle and reliability evaluation index;

[0050] The evaluation indicators corresponding to the grid-connected performance evaluation angle include an inertia response evaluation indicator, a primary frequency regulation evaluation indicator, an automatic generation control evaluation indicator, and an automatic voltage control evaluation indicator;

[0051] The evaluation indicators corresponding to the reliability evaluation angle include battery reliability evaluation indicators and power station reliability evaluation indicators; the battery reliability evaluation indicators include battery consistency indicators, battery health status and battery failure indicators, and the power station reliability evaluation indicators include power station outage coefficient, power station availability coefficient and power station utilization coefficient.

[0052] Optionally, the dimension reduction module includes:

[0053] A first standardization submodule is used to perform standardization processing on the original matrix to obtain a standardized matrix;

[0054] A feature screening submodule, used to obtain a corresponding correlation coefficient matrix based on the correlation coefficients between the elements in the standardized matrix, and screen out a specified number of target eigenvalues ​​based on the eigenvalues ​​of the correlation coefficient matrix;

[0055] A eigenvector determination submodule, used to determine a corresponding target eigenvector based on the target eigenvalue to obtain an eigenvector matrix;

[0056] The dimension reduction matrix determination submodule is used to obtain the dimension reduction matrix based on the standardized matrix and the eigenvector matrix.

[0057] Optionally, the first standardization submodule is specifically used for:

[0058] Standardize each element of the original matrix based on the mean and standard deviation of each evaluation index in the original matrix to obtain a standardized matrix;

[0059] The dimension reduction matrix determination submodule is specifically used for:

[0060] For each column of data in the standardized matrix, each element in each column of data is normalized based on the index mean of each column of data to obtain a sample matrix;

[0061] The product of the sample matrix and the eigenvector matrix is ​​used as the dimension reduction matrix.

[0062] Optionally, the weight determination module includes:

[0063] A weight scoring submodule, used for scoring the weight of each evaluation index in the dimensionality reduction matrix by each expert, obtaining an initial expert weight matrix corresponding to the dimensionality reduction matrix, and using each column of the initial expert weight matrix as a comparison data column;

[0064] A screening submodule, used for screening out the maximum value of each column from the initial expert weight matrix to form a reference data column;

[0065] The indicator weight determination submodule is used to determine the indicator weight of the evaluation indicator corresponding to each comparison data column based on the Mahalanobis distance between each comparison data column and the reference data column.

[0066] Optionally, the calculation formula of the indicator weight of the evaluation indicator is:

[0067]

[0068] Among them, ω k represents the index weight of the kth evaluation index, ω′ k Represents ω k The data before normalization, E0 represents the reference data column, E k represents the kth comparison data column, k=1, 2, ..., p, and p is the number of evaluation indicators in the dimension reduction matrix; D s represents the Mahalanobis distance.

[0069] Optionally, the evaluation module includes:

[0070] A second standardization submodule is used to perform maximum-minimum standardization processing on each element of the dimension reduction matrix to obtain a standardized dimension reduction matrix;

[0071] The clustering submodule is used to use the row vectors of the standardized dimensionality reduction matrix as data points, the distance between two row vectors as the distance between data points, and use a density-based clustering algorithm to cluster the row vectors in the standardized dimensionality reduction matrix to obtain the center point and coordinates of each cluster cluster; the center point of the cluster cluster is used to indicate the average score of each energy storage power station in the cluster cluster, and the coordinates of the center point in each evaluation index dimension are used to indicate the average score of all energy storage power stations in the corresponding cluster cluster for each evaluation index.

[0072] Optionally, the evaluation module includes:

[0073] The score calculation submodule is used to accumulate the product of the coordinates of the center point of each cluster in each evaluation index dimension and the corresponding index weight for each cluster, so as to obtain a comprehensive score for each cluster;

[0074] The evaluation submodule is used to evaluate each energy storage power station in the cluster based on the comprehensive score of each cluster.

[0075] On the other hand, the present invention also provides an electronic device, comprising: at least one processor and a memory; the memory and the processor are connected via a bus;

[0076] The memory is used to store one or more programs;

[0077] When the one or more programs are executed by the at least one processor, any one of the above-mentioned large-scale energy storage power station evaluation methods based on multi-dimensional data fusion is implemented.

[0078] On the other hand, the present invention also provides a readable storage medium having an execution program stored thereon, and when the execution program is executed, any one of the above-mentioned large-scale energy storage power station evaluation methods based on multi-dimensional data fusion is implemented.

[0079] Compared with the prior art, the present invention has the following beneficial effects:

[0080] The present invention provides a method and system for evaluating a large-scale energy storage power station based on multi-dimensional data fusion. On the one hand, the values ​​of each evaluation index corresponding to multiple evaluation angles are calculated based on the multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station, and a multi-angle evaluation index system based on multi-dimensional data is constructed. Based on the energy storage power station, the values ​​of each evaluation index corresponding to multiple evaluation angles are fused into the original matrix of the evaluation index of the large-scale energy storage power station. Based on the multi-dimensional and multi-angle data fusion, a comprehensive evaluation of the large-scale energy storage power station is carried out in an overall manner, thereby improving the accuracy of the evaluation and grading of the large-scale energy storage power station. On the other hand, the original matrix is ​​reduced in dimension by using the principal component analysis method, thereby reducing the data analysis dimension and minimizing the information loss as much as possible; the expert weights of each evaluation index in the reduced dimension matrix are correlated by using the grey correlation method to measure the correlation between the index data sets, thereby avoiding the influence of different dimensions and resolution coefficient values, reducing the subjective uncertainty of the evaluation, and further improving the accuracy of the evaluation, thereby ensuring the rationality of the operation and maintenance decision of the large-scale energy storage power station based on the evaluation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 It is a flow chart of a method for evaluating a large-scale energy storage power station based on multi-dimensional data fusion according to the present invention;

[0082] Figure 2 It is a schematic diagram of an evaluation index system based on multi-dimensional data fusion of the present invention;

[0083] Figure 3 It is a schematic diagram of a process of performing dimensionality reduction processing on the original matrix by using the principal component analysis method of the present invention;

[0084] Figure 4 It is a flow chart of the present invention that uses the grey correlation method to perform correlation processing on the expert weights of the evaluation indicators in the dimension reduction matrix;

[0085] Figure 5 It is a schematic structural diagram of the electronic device of the present invention. DETAILED DESCRIPTION

[0086] The specific implementation modes of the present invention are further described in detail below with reference to the accompanying drawings.

[0087] Embodiment 1:

[0088] The present invention provides a method for evaluating a large-scale energy storage power station based on multi-dimensional data fusion, as shown in the schematic diagram Figure 1 As shown, including:

[0089] Step S110, calculating the values ​​of each evaluation index corresponding to multiple evaluation angles based on the acquired multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station, and fusing the values ​​of each evaluation index corresponding to the multiple evaluation angles into an original matrix of the evaluation index of the large-scale energy storage power station based on the energy storage power station;

[0090] Step S120, using principal component analysis to perform dimensionality reduction processing on the original matrix to obtain a corresponding dimensionality reduction matrix;

[0091] Step S130, using the grey correlation method to correlate the expert weights of the evaluation indicators in the dimension reduction matrix to obtain the indicator weights of the evaluation indicators;

[0092] Step S140 , clustering the values ​​of the evaluation indicators in the dimension reduction matrix, and evaluating the energy storage power stations in each cluster based on the values ​​of the evaluation indicators in each cluster in the clustering result and the corresponding indicator weights.

[0093] In this example implementation, the multi-dimensional operation data may include data from the battery management system (BMS) collected by the energy storage power station, data from the power conversion system (PCS), and dispatch telemetry data. Figure 2As shown, multiple evaluation angles may include energy efficiency evaluation angle, regulation performance evaluation angle, grid-connected performance evaluation angle and reliability evaluation index; the evaluation indexes corresponding to the energy efficiency evaluation angle include utilization hours, utilization rate, cycle efficiency, number of charge and discharge times and equivalent cycle coefficient, etc.; the evaluation indexes corresponding to the regulation performance angle include regulation speed, regulation accuracy and response time, etc.; the evaluation indexes corresponding to the grid-connected performance evaluation angle include inertia response evaluation index, primary frequency regulation evaluation index, automatic generation control (AGC) evaluation index and automatic voltage control (AVC) evaluation index; the evaluation indexes corresponding to the reliability evaluation angle include battery reliability evaluation index and power station reliability evaluation index; the battery reliability evaluation index includes battery consistency index, battery health status, namely battery SOH (State of Health) and battery failure index (battery or battery stack failure rate), and the power station reliability evaluation index includes power station outage coefficient (power station planned outage coefficient and power station unplanned outage coefficient), power station availability coefficient (power station equivalent availability coefficient) and power station utilization coefficient (power station equivalent utilization coefficient). Among them, the inertia response evaluation indicators include the inertia response operation rate and the inertia response output contribution; the primary frequency regulation evaluation includes the primary frequency regulation effect, the primary frequency regulation operation rate and the primary frequency regulation output contribution; the AGC evaluation indicators include the AGC operation rate and the AGC output contribution; the AVC evaluation indicators include the AVC operation rate and the AVC regulation pass rate; the battery consistency indicators include the battery terminal voltage consistency, the battery capacity consistency and the battery internal resistance consistency.

[0094] For example, the calculation method of each evaluation index of the energy storage power station can refer to the current industry standards or specifications, and the specific calculation method is as follows:

[0095] (1) Evaluation indicators corresponding to the energy efficiency evaluation perspective:

[0096] Cycle efficiency refers to the ratio of the energy output of the energy storage station during rated power discharge to the energy input during rated power charging in the same cycle. The calculation formula is:

[0097]

[0098] Wherein, η represents the cycle efficiency; the energy storage station is first discharged at rated power to the discharge termination condition and then stopped discharging, and then charged at rated power to the charging termination condition and then stopped charging. In this process, the energy charged by the energy storage station during the charging process is b, and the energy discharged by the energy storage station during the discharging process is a. The number of utilization hours refers to the number of operating hours of the energy storage station at rated power during the evaluation period. The utilization rate refers to the ratio of the number of utilization hours to the evaluation period. The number of charge and discharge times refers to the count value obtained by dividing the cumulative charge and discharge capacity of the energy storage station under various operating conditions by the nominal rated capacity of the station. The calculation formula for the equivalent cycle coefficient is:

[0099]

[0100] In the formula, ε represents the equivalent cycle coefficient, d represents the number of days of the evaluation cycle, Q n represents the rated capacity of the battery, a1 represents the cumulative discharge capacity during the evaluation period, and b1 represents the cumulative charge capacity during the evaluation period.

[0101] (2) The regulation speed in the evaluation index corresponding to the regulation performance evaluation angle refers to the speed at which the energy storage power station responds to the frequency regulation command. Its calculation formula is:

[0102]

[0103] Among them, v i is the regulation speed of the i-th energy storage power station; are the regulation command value of the i-th energy storage power station and the initial power value when participating in regulation; Δt i is the adjustment time of the i-th energy storage power station; v N Adjust the speed to the standard; is the regulation speed performance of the i-th energy storage power station. The regulation accuracy refers to the difference between the final actual output of the energy storage power station and the regulation command value. The smaller the difference, the better the performance. The calculation formula is:

[0104]

[0105] Among them, ΔP i is the regulation error of the i-th energy storage power station, t is the time; ΔP N is the standard adjustment error; is the regulation accuracy performance of the i-th energy storage power station.

[0106] Response time refers to the time required for the energy storage power station to start adjusting its output from receiving the adjustment command. The smaller the response time, the better the response performance. The calculation formula is:

[0107]

[0108] Among them, ti is the response time of the i-th energy storage power station; t N is the standard response time; is the response time performance of the i-th energy storage power station.

[0109] (3) The calculation formula of the inertia response evaluation index in the evaluation index corresponding to the grid-connected performance is:

[0110]

[0111] Among them, λ represents the operation rate of inertia response, t1 represents the operation time of inertia response, t2 represents the evaluation cycle time, α represents the output contribution of inertia response, P1 represents the output at the end of inertia response, and P2 represents the output at the beginning of inertia response.

[0112] The calculation formula of primary frequency modulation evaluation index is:

[0113]

[0114] γ=P3-P4

[0115]

[0116] Among them, β represents the commissioning rate of primary frequency regulation; t3 represents the commissioning time of primary frequency regulation within the evaluation period; t2 represents the evaluation period time; γ represents the output contribution of primary frequency regulation; P3 represents the power at the end of primary frequency regulation; P4 represents the power at the beginning of primary frequency regulation; DX is the performance index of primary frequency regulation effect; ΔQsY is the actual action integral power of primary frequency regulation, ΔQjY is the theoretical calculation integral power of primary frequency regulation during response time; t0 is the moment when the grid frequency is equal to 50±ΔfsqHz, t 调节 is the time when the grid frequency exceeds 50±ΔfsqHz, Δfsq is the frequency control dead zone, and the Δfsq value of the electric energy storage is 0.05Hz; t 调节 The maximum value is 60 seconds; PST is the actual charging and discharging power of the energy storage at time t0; PSt is the time after t0 调节 The actual charging and discharging power of the energy storage within a certain period of time. The AGC evaluation indicators include the AGC operation rate and the AGC output contribution; the AVC evaluation indicators include the AVC operation rate and the AVC regulation qualification rate. Δf is the value of the grid frequency exceeding 50±Δfsq, with high frequency as a positive value and low frequency as a negative value. ΔP(Δf,t) is the theoretical response output change.

[0117] (4) The battery terminal voltage consistency in the battery reliability evaluation index refers to the terminal voltage difference between each battery in the battery pack, and its calculation formula is:

[0118]

[0119] Among them, ε U is the battery terminal voltage consistency; U i is the terminal voltage of the i-th single cell; is the average terminal voltage of all single cells; ε′ U is the mean square error of the battery terminal voltage, and M is the number of single cells. Battery capacity consistency refers to the capacity difference between each battery in the battery pack. Battery internal resistance consistency refers to the internal resistance difference between each battery in the battery pack. Battery SOH refers to the health status and performance degradation of the battery, and its calculation formula is:

[0120]

[0121] Among them, Q n Indicates the rated capacity of the battery; Q d Indicates the actual capacity of the battery. The battery (stack) cluster failure rate refers to the ratio of the number of battery (stack) cluster failures in the energy storage unit to the total number of battery (stack) clusters.

[0122] (5) The planned outage coefficient of the power station in the power station reliability evaluation index refers to the ratio of the planned outage time of the energy storage power station to the statistical time within the evaluation period. Its calculation formula is:

[0123]

[0124] Among them, POF is the planned outage coefficient of the energy storage power station (%); POH is the planned outage hours in the evaluation period, in hours; PH is the statistical time hours in the evaluation period, in hours. When the evaluation period is 1 year, the PH value is 8760h. The unplanned outage coefficient of the power station refers to the ratio of the unplanned outage time of the energy storage power station to the statistical time in the evaluation period. The availability coefficient of the energy storage power station refers to the ratio of the available time of the power station to the statistical time in the evaluation period. The utilization coefficient of the energy storage power station refers to the ratio of the operating time of the power station to the statistical time in the evaluation period.

[0125] Through the above process, we obtain the evaluation indicators of large-scale energy storage power stations from multiple angles, and use the energy storage power station as the row vector dimension to integrate the evaluation indicators, and construct the original matrix of the evaluation indicators of large-scale energy storage power stations. For example, for the n evaluation indicators of m energy storage power stations, we construct an m×n energy storage power station-evaluation indicator matrix as the original matrix A. For example, if 1,000 energy storage power stations need to be evaluated, the original indicator data consisting of 26 evaluation indicators need to be obtained from the master station for each energy storage power station, forming a 1,000×26 original indicator data matrix A. This example conducts a comprehensive evaluation of energy storage power stations from multiple angles such as energy efficiency, regulation capability, grid-connected performance, and reliability, and sorts out the relevant data.

[0126] In an example implementation, before calculating the values ​​of each evaluation index corresponding to multiple evaluation angles based on the acquired multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station in step S110, the method further includes the following process:

[0127] Collecting the energy dimension operation data of each energy storage power station in the large-scale energy storage power station through the energy management system of the large-scale energy storage power station;

[0128] Collecting the operation data of the power regulation dimension of each energy storage power station in the large-scale energy storage power station through the energy storage converter of the large-scale energy storage power station;

[0129] The battery management system of the large-scale energy storage power station is used to collect the operating data of the battery dimension of each energy storage power station in the large-scale energy storage power station.

[0130] In this example implementation, multi-dimensional operation data can be obtained by collecting the operation data of each energy storage power station by the energy management system, energy storage converter and battery management system in the energy storage power station. Specifically, the data collected by the energy management system include inertia response commissioning time, inertia response commissioning evaluation cycle time, primary frequency regulation commissioning time, primary frequency regulation commissioning evaluation cycle time, AGC commissioning time, AGC commissioning evaluation cycle time, AGC each adjustment power, AGC each adjustment time, AVC commissioning time, AVC commissioning evaluation cycle time, AVC qualified adjustment times, AVC total adjustment times, actual discharge capacity of the power station, actual current capacity of the battery, energy storage power station adjustment command value, adjustment time, response time, time t0 when the grid frequency is equal to 50±ΔfsqHz, time t when the grid frequency exceeds 50±ΔfsqHz 调节 , planned outage hours during the evaluation period, unplanned outage hours during the evaluation period, available hours during the evaluation period, power station operation hours during the evaluation period, and statistical time hours during the evaluation period. The data collected by the energy storage inverter include the maximum allowable discharge power, the maximum allowable charging power, the power at the start of inertia response, the power at the end of inertia response, the power at the start of primary frequency regulation, the power at the end of primary frequency regulation, the power at the start of AGC, the power at the end of AGC, the initial power value when the energy storage power station participates in regulation, the actual output average value of the generator set within 10 seconds before time t0 or the actual charging and discharging power of the energy storage at time t0, the power of the generator set after time t0, and the actual power consumption of the generator set within 10 seconds before time t0. 调节The actual output of the unit or the actual charging and discharging power of the electric energy storage within the time, etc. The data collected by the battery management system include charging voltage, charging current, discharging voltage, discharging current, cumulative charging capacity, cumulative discharging capacity, rated capacity, battery temperature, terminal voltage of single cell, capacitance of single cell, internal resistance of single cell, relative failure times of energy storage unit battery or battery stack, failure times of battery or battery stack, and total number of batteries or battery stacks in the unit, etc. By collecting the operating data of the above multiple dimensions and using the calculation formulas of various evaluation indicators of the evaluation indicator system, the original value of each evaluation indicator in the constructed evaluation indicator system is calculated.

[0131] In an example implementation, the step S120 of performing dimensionality reduction processing on the original matrix using the principal component analysis method to obtain a corresponding dimensionality reduction matrix includes:

[0132] Performing normalization processing on the original matrix to obtain a normalized matrix;

[0133] Obtaining a corresponding correlation coefficient matrix based on the correlation coefficients between the elements in the standardized matrix, and selecting a specified number of target eigenvalues ​​based on the eigenvalues ​​of the correlation coefficient matrix;

[0134] Determine a corresponding target eigenvector based on the target eigenvalue to obtain an eigenvector matrix;

[0135] The dimension reduction matrix is ​​obtained based on the normalized matrix and the eigenvector matrix.

[0136] In this example implementation, the original matrix is ​​first normalized to reduce the amount of calculation. Specifically, Figure 3 As shown, each element of the original matrix can be standardized based on the mean and standard deviation of each evaluation index in the original matrix to obtain a standardized matrix; that is, the original indicator data matrix is ​​standardized, and the standardized processing is performed according to the following formula:

[0137]

[0138] Among them, a ij is the element value of the jth index in the i-th row of the original matrix A, i.e., the original index data, a′ ij For a ij The corresponding standardized data, x j and b j are the mean and standard deviation of the jth indicator respectively. Then, the correlation coefficient matrix R is calculated based on the standardized matrix. Specifically, the correlation coefficients between the elements in the standardized matrix are calculated, and the correlation coefficient matrix R is composed of the correlation coefficients, which is specifically expressed as follows:

[0139]

[0140] Among them, r xy = Cov(a′ x ,a′ y ), Cov is the covariance, r xy is the xth element a′ of the normalized matrix x and the yth element a′ y The correlation coefficient of x and y is 1, 2, ..., m*n, r respectively. xy =r yx After obtaining the correlation coefficient matrix R, the eigenvalues ​​and eigenvectors of the correlation coefficient matrix R are calculated. For example, 26 eigenvalues ​​and 26 eigenvectors are obtained. Then, the eigenvalues ​​are sorted in descending order, and the first p eigenvalues ​​are selected as the target eigenvalues. When selecting the target eigenvalue, the sum of the first p eigenvalues ​​(target eigenvalues) must be at least greater than 90% of the sum of all 26 eigenvalues, and the eigenvectors corresponding to the target eigenvalues ​​are screened out to form an eigenvector matrix C (with a dimension of 26×p). Exemplarily, for each column of the standardized matrix, each element in each column of the data is normalized based on the mean of the index of each column of the data to obtain a sample matrix; that is, the mean of the index data of each column is calculated according to the standardized matrix, and then each element of the standardized matrix is ​​subtracted from the mean of the index of the corresponding column to obtain a sample matrix B (with a dimension of 1000×26); the product of the sample matrix and the eigenvector matrix is ​​used as the dimension reduction matrix, that is, the sample matrix B is multiplied by the eigenvector matrix C to obtain the dimension reduction matrix D (with a dimension of 1000×p) after dimension reduction. At this time, the number of evaluation indicators is reduced from 26 to p, p<26, and the dimension reduction of the evaluation indicators is achieved. This example uses the principal component analysis method to reduce the dimension of the multidimensional massive data of the energy storage power station. While reducing the dimension of data analysis, it minimizes the loss of information contained in the original dimension to achieve the purpose of comprehensive analysis of the collected data.

[0141] In some example implementations, the use of the grey correlation method in step S130 to correlate the initial expert weights of the evaluation indicators in the dimensionality reduction matrix to obtain the indicator weights of the evaluation indicators includes the following process:

[0142] Based on the weight score of each expert for each evaluation indicator in the dimensionality reduction matrix, an initial expert weight matrix corresponding to the dimensionality reduction matrix is ​​obtained, and each column of the initial expert weight matrix is ​​used as a comparison data column;

[0143] Screening out the maximum value of each column from the initial expert weight matrix to form a reference data column;

[0144] Based on the Mahalanobis distance between each comparison data column and the reference data column, an indicator weight of the evaluation indicator corresponding to each comparison data column is determined.

[0145] In this example implementation, the process of determining the indicator weight is as follows: Figure 4 As shown, first, q experts are set to make empirical judgments on the weights of p evaluation indicators at the same time. Each expert must give a weighted score to each evaluation indicator to form an initial expert weight matrix E as a comparison data column, as shown in the following formula:

[0146] E=[E1,E2,…,E p ] T ;

[0147]

[0148] The superscript T represents the transpose of the matrix, E p is the index sequence of the pth evaluation index, a 1q is the weighted score of the qth expert on the first evaluation indicator, a p1 is the weighted score of the first expert for the pth evaluation index, a pq is the weighted score given by the qth expert to the pth evaluation indicator.

[0149] Secondly, select the highest weight value, i.e. the maximum value, of each column from the comparison data column E and combine them into the reference data column E0:

[0150] E0=[a 01 ,a 02 ,…,a 0q ];

[0151] Among them, a 01 represents the maximum value of the first column of the initial expert weight matrix E, a 0q Represents the maximum value of the qth column of the initial expert weight matrix E. The reference data column E0 and the comparison data column E can be synthesized into a new matrix E' to calculate the Mahalanobis distance between the two, as shown below:

[0152] E'=[E0,E1,E2,…,E p ] T ;

[0153]

[0154] Then, calculate each comparison data column E1, E2, ..., E p The Mahalanobis distance D with the reference data column E0 s :

[0155]

[0156] Where, k = 1, 2, ..., p; ∑ is the covariance matrix of the matrix E'; D s (E0,E k ) is the reference data column E0 and the kth comparison data column E k After obtaining the Mahalanobis distance between each comparison data column and the reference data column, finally, the weight of each indicator is calculated and normalized. The calculation formula of the indicator weight of the specific evaluation indicator is:

[0157]

[0158] Among them, ω k represents the index weight of the kth evaluation index, ω′ k Represents ω k The data before normalization, E0 represents the reference data column, E k represents the kth comparison data column, k=1, 2, ..., p, and p is the number of evaluation indicators in the dimension reduction matrix; D s Represents Mahalanobis distance. This example uses the improved grey correlation method to measure the correlation between indicator data sets, avoiding the influence of the value of the resolution coefficient ρ and reducing the subjective uncertainty of the evaluation. At the same time, it changes the way of quantifying the similarity between the comparison data column and the reference data column, and proposes to use Mahalanobis distance to quantify the similarity between the two. Mahalanobis distance is not only not affected by the dimension, but also more suitable for measuring the similarity between two sets of vectors than Euclidean distance. This example uses Mahalanobis distance to correct Euclidean distance, and corrects the related problems of inconsistent scales of various dimensions caused by using Euclidean distance calculation.

[0159] In some implementations, the step S140 of clustering the values ​​of the evaluation indicators in the dimension reduction matrix includes:

[0160] Performing maximum-minimum normalization processing on each element of the dimension reduction matrix to obtain a standardized dimension reduction matrix;

[0161] The row vectors of the standardized dimensionality reduction matrix are used as data points, and the distance between two row vectors is used as the distance between data points. The row vectors in the standardized dimensionality reduction matrix are clustered using a density-based clustering algorithm to obtain the center point of each cluster cluster and its coordinates; the center point of the cluster cluster is used to indicate the average score of each energy storage power station in the cluster cluster, and the coordinate of the center point in each evaluation index dimension is used to indicate the average score of all energy storage power stations in the corresponding cluster cluster for each evaluation index.

[0162] In this example implementation, in order to simplify the clustering calculation process, each data in the dimension reduction matrix D is subjected to maximum-minimum normalization, and the data is mapped to the range [0, 1]. The calculation formula is:

[0163]

[0164] Among them, a uv represents the value of the index (column) of the uth row of the dimension reduction matrix D; a vmin represents the minimum value of the vth index of the dimension reduction matrix D; a vmax represents the maximum value of the vth index of the dimension reduction matrix D; a′ uv Represents the value of the vth indicator (column) in the uth row of the standardized dimension reduction matrix. After standardization, the DBSCAN clustering method is used to cluster the energy storage power station. The specific process is: select an appropriate neighborhood radius Eps and a threshold value of the number of data objects in the neighborhood MinPts. Since the data is mapped to [0,1] during standardization, the domain radius Eps should not exceed 0.4, otherwise there will be too few clusters. The specific domain radius and object number threshold can be determined based on the desired number of evaluation levels. For example, suppose there are 1,000 objects to be evaluated (energy storage power stations) and they are to be divided into 4 evaluation levels. Then the domain radius can be set to about 0.2, and the object number threshold can be set between [30,60]. The settings of these two parameters can be fine-tuned according to the characteristics of the data set. Cluster the indicator data set (standardized dimensionality reduction matrix) according to the above parameter settings. Specifically, the row vectors of the standardized dimensionality reduction matrix are used as data points, and the distance between two row vectors is used as the distance between data points. Cluster clusters are searched based on the domain radius Eps. If the points contained in the neighborhood radius Eps of any data point exceed the object number threshold MinPts, a cluster with the data point as the core object is created. Objects that can be reached from the core object density are iteratively clustered until no new points are added to any cluster. The clustering ends and the center point and coordinates of each cluster are obtained. The cluster center point of the cluster represents the overall average level of the energy storage power stations in this cluster. The coordinates of the center point include multiple indicator dimension values. The coordinates on each indicator dimension represent the average score of all energy storage power stations in the corresponding cluster under each indicator.

[0165] In some example implementations, the step S140 of evaluating each energy storage power station in each cluster based on the value of each evaluation indicator in each cluster in the clustering result and the corresponding indicator weight includes:

[0166] For each cluster, the product of the coordinates of the center point of the cluster in each evaluation index dimension and the corresponding index weight is accumulated to obtain a comprehensive score for each cluster;

[0167] The energy storage power stations within each cluster are evaluated based on the comprehensive score of each cluster.

[0168] In this example implementation, the coordinate components of the center point of each cluster, i.e., the index value under each evaluation index dimension, are multiplied by the corresponding index weight and then summed to obtain the comprehensive score of each cluster center point. Finally, the score values ​​are sorted from large to small. Each cluster center point corresponds to a cluster cluster of energy storage power station index data set, so this order is also the order of cluster clusters, and it is also the overall average score order of all energy storage power stations in each cluster. According to this order, the evaluation level of energy storage power stations can be divided, and the classification evaluation of large-scale energy storage power stations can be realized. Based on the comprehensive score of the energy storage power station, a comprehensive evaluation of the large-scale energy storage power station is carried out. The larger the comprehensive score, the higher the rating of the energy storage power station. The operation and maintenance decision of the large-scale energy storage power station can be guided based on the evaluation level of the energy storage power station. The present invention can improve the accuracy of the evaluation grading and the rationality of the operation and maintenance decision.

[0169] The present invention is aimed at large-scale energy storage power stations, which refer to a large number of energy storage power stations. Due to the large number of energy storage power stations, the evaluation process of energy storage power stations has the characteristics of large data volume and numerous attributes, and it is difficult to conduct a comprehensive evaluation using traditional evaluation methods. The present invention designs an energy storage power station evaluation index system from multiple angles such as energy efficiency, regulation capability, grid-connected performance, and reliability to achieve a comprehensive analysis of large-scale energy storage power stations; the principal component analysis method is used to pre-process the multi-dimensional massive data of large-scale energy storage power stations, while reducing the dimension of data analysis, minimizing the loss of information contained in the original dimension, so as to achieve the purpose of comprehensive analysis of the collected data; the improved grey correlation method is used to calculate the index weight to measure the correlation between the index data sets, which can avoid the influence of the dimension, and at the same time avoid the influence of the resolution coefficient value, reducing the subjective uncertainty of the evaluation; the DBSCAN clustering method is used to divide the evaluation level of the energy storage power station, and a classified and comprehensive evaluation of large-scale energy storage power stations is achieved. Based on data fusion technology, comprehensive evaluation of energy storage power stations can be carried out, and multiple large-scale energy storage power stations can be evaluated at the same time, which improves the accuracy of evaluation classification and the rationality of operation and maintenance decisions, and at the same time improves the evaluation efficiency of energy storage power stations.

[0170] It should be noted that the present invention is not only applicable to the evaluation of large-scale energy storage power stations, but also applicable to the evaluation of energy storage power stations that are small in number and have not yet formed a scale.

[0171] Embodiment 2:

[0172] Based on the same inventive concept, the present invention also provides a large-scale energy storage power station evaluation system based on multi-dimensional data fusion, the system comprising:

[0173] A data fusion module is used to calculate the values ​​of each evaluation index corresponding to multiple evaluation angles based on the multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station obtained, and fuse the values ​​of each evaluation index corresponding to the multiple evaluation angles into an original matrix of the evaluation index of the large-scale energy storage power station based on the energy storage power station;

[0174] A dimension reduction module is used to perform dimension reduction processing on the original matrix using a principal component analysis method to obtain a corresponding dimension reduction matrix;

[0175] A weight determination module is used to perform correlation processing on the expert weights of each evaluation index in the dimension reduction matrix using a grey correlation method to obtain an index weight of each evaluation index;

[0176] The evaluation module is used to cluster the values ​​of the evaluation indicators in the dimension reduction matrix, and evaluate the energy storage power stations in each cluster based on the values ​​of the evaluation indicators in each cluster and the corresponding indicator weights in the cluster results.

[0177] In a possible implementation, the system further includes a data acquisition module, and the data acquisition module is used to:

[0178] Collecting the energy dimension operation data of each energy storage power station in the large-scale energy storage power station through the energy management system of the large-scale energy storage power station;

[0179] Collecting the operation data of the power regulation dimension of each energy storage power station in the large-scale energy storage power station through the energy storage converter of the large-scale energy storage power station;

[0180] The battery management system of the large-scale energy storage power station is used to collect the operating data of the battery dimension of each energy storage power station in the large-scale energy storage power station.

[0181] In a possible implementation manner, the evaluation angles include energy efficiency evaluation angle, regulation performance evaluation angle, grid connection performance evaluation angle and reliability evaluation index;

[0182] The evaluation indicators corresponding to the grid-connected performance evaluation angle include an inertia response evaluation indicator, a primary frequency regulation evaluation indicator, an automatic generation control evaluation indicator, and an automatic voltage control evaluation indicator;

[0183] The evaluation indicators corresponding to the reliability evaluation angle include battery reliability evaluation indicators and power station reliability evaluation indicators; the battery reliability evaluation indicators include battery consistency indicators, battery health status and battery failure indicators, and the power station reliability evaluation indicators include power station outage coefficient, power station availability coefficient and power station utilization coefficient.

[0184] In a possible implementation, the dimension reduction module includes:

[0185] A first standardization submodule is used to perform standardization processing on the original matrix to obtain a standardized matrix;

[0186] A feature screening submodule, used to obtain a corresponding correlation coefficient matrix based on the correlation coefficients between the elements in the standardized matrix, and screen out a specified number of target eigenvalues ​​based on the eigenvalues ​​of the correlation coefficient matrix;

[0187] A eigenvector determination submodule, used to determine a corresponding target eigenvector based on the target eigenvalue to obtain an eigenvector matrix;

[0188] The dimension reduction matrix determination submodule is used to obtain the dimension reduction matrix based on the standardized matrix and the eigenvector matrix.

[0189] In a possible implementation manner, the first standardization submodule is specifically configured to:

[0190] Standardize each element of the original matrix based on the mean and standard deviation of each evaluation index in the original matrix to obtain a standardized matrix;

[0191] The dimension reduction matrix determination submodule is specifically used for:

[0192] For each column of data in the standardized matrix, each element in each column of data is normalized based on the index mean of each column of data to obtain a sample matrix;

[0193] The product of the sample matrix and the eigenvector matrix is ​​used as the dimension reduction matrix.

[0194] In a possible implementation, the weight determination module includes:

[0195] A weight scoring submodule, used for scoring the weight of each evaluation index in the dimensionality reduction matrix by each expert, obtaining an initial expert weight matrix corresponding to the dimensionality reduction matrix, and using each column of the initial expert weight matrix as a comparison data column;

[0196] A screening submodule, used for screening out the maximum value of each column from the initial expert weight matrix to form a reference data column;

[0197] The indicator weight determination submodule is used to determine the indicator weight of the evaluation indicator corresponding to each comparison data column based on the Mahalanobis distance between each comparison data column and the reference data column.

[0198] In a possible implementation, the calculation formula of the indicator weight of the evaluation indicator is:

[0199]

[0200] Among them, ωk represents the index weight of the kth evaluation index, ω′ k Represents ω k The data before normalization, E0 represents the reference data column, E k represents the kth comparison data column, k=1, 2, ..., p, and p is the number of evaluation indicators in the dimension reduction matrix; D s represents the Mahalanobis distance.

[0201] In a possible implementation, the evaluation module includes:

[0202] A second standardization submodule is used to perform maximum-minimum standardization processing on each element of the dimension reduction matrix to obtain a standardized dimension reduction matrix;

[0203] The clustering submodule is used to use the row vectors of the standardized dimensionality reduction matrix as data points, the distance between two row vectors as the distance between data points, and use a density-based clustering algorithm to cluster the row vectors in the standardized dimensionality reduction matrix to obtain the center point and coordinates of each cluster cluster; the center point of the cluster cluster is used to indicate the average score of each energy storage power station in the cluster cluster, and the coordinates of the center point in each evaluation index dimension are used to indicate the average score of all energy storage power stations in the corresponding cluster cluster for each evaluation index.

[0204] In a possible implementation, the evaluation module includes:

[0205] The score calculation submodule is used to accumulate the product of the coordinates of the center point of each cluster in each evaluation index dimension and the corresponding index weight for each cluster, so as to obtain a comprehensive score for each cluster;

[0206] The evaluation submodule is used to evaluate each energy storage power station in the cluster based on the comprehensive score of each cluster.

[0207] Embodiment 3:

[0208] like Figure 5 As shown, the present invention also provides an electronic device, which may be a computer device, a single-chip device, an intelligent mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, the processor, and the transceiver component are connected via a bus; the memory may be used to store an execution program, and an exemplary execution program may include instructions; the processor is used to execute the instructions stored in the memory. The memory may also be used to store data, which may be called and / or modified when the instructions are executed.

[0209] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions in a storage medium to implement the corresponding method flow or corresponding function, so as to implement the steps of a large-scale energy storage power station evaluation method based on multi-dimensional data fusion in the above embodiment.

[0210] Example 4

[0211] Based on the same inventive concept, the present invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory), which is a memory device in an electronic device for storing programs and data. It can be understood that the storage medium here can include both built-in storage media in electronic devices and, of course, extended storage media supported by electronic devices. The storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more execution programs (including program codes). It should be noted that the storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor loads and executes one or more instructions stored in the storage medium, which can implement the steps of a large-scale energy storage power station evaluation method based on multi-dimensional data fusion in the above embodiment.

[0212] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0213] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0214] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0215] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0216] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that after reading the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the application, but these changes, modifications or equivalent substitutions are all within the protection scope of the claims to be approved.

Claims

1. A large-scale energy storage power station evaluation method based on multi-dimensional data fusion, characterized in that: include: Calculate the values ​​of each evaluation index corresponding to multiple evaluation angles based on the acquired multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station, and fuse the values ​​of each evaluation index corresponding to the multiple evaluation angles into an original matrix of the evaluation index of the large-scale energy storage power station based on the energy storage power station; Performing dimensionality reduction processing on the original matrix using principal component analysis to obtain a corresponding dimensionality reduction matrix; The expert weights of the evaluation indicators in the dimension reduction matrix are correlated by using the grey correlation method to obtain the indicator weights of the evaluation indicators; The values ​​of the evaluation indicators in the dimension reduction matrix are clustered, and the energy storage power stations in each cluster are evaluated based on the values ​​of the evaluation indicators in each cluster in the clustering result and the corresponding indicator weights.

2. The method according to claim 1, characterized in that Before calculating the values ​​of each evaluation index corresponding to multiple evaluation angles based on the acquired multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station, the method further includes: Collecting the energy dimension operation data of each energy storage power station in the large-scale energy storage power station through the energy management system of the large-scale energy storage power station; Collecting the operation data of the power regulation dimension of each energy storage power station in the large-scale energy storage power station through the energy storage converter of the large-scale energy storage power station; The battery management system of the large-scale energy storage power station is used to collect the operating data of the battery dimension of each energy storage power station in the large-scale energy storage power station.

3. The method according to claim 2, characterized in that The evaluation angles include energy efficiency evaluation angle, regulation performance evaluation angle, grid connection performance evaluation angle and reliability evaluation index; The evaluation indicators corresponding to the grid-connected performance evaluation angle include an inertia response evaluation indicator, a primary frequency regulation evaluation indicator, an automatic generation control evaluation indicator, and an automatic voltage control evaluation indicator; The evaluation indicators corresponding to the reliability evaluation angle include battery reliability evaluation indicators and power station reliability evaluation indicators; the battery reliability evaluation indicators include battery consistency indicators, battery health status and battery failure indicators, and the power station reliability evaluation indicators include power station outage coefficient, power station availability coefficient and power station utilization coefficient.

4. The method according to claim 1, characterized in that: The method of reducing the dimension of the original matrix by using the principal component analysis method to obtain a corresponding reduced dimension matrix includes: Performing normalization processing on the original matrix to obtain a normalized matrix; Obtaining a corresponding correlation coefficient matrix based on the correlation coefficients between the elements in the standardized matrix, and selecting a specified number of target eigenvalues ​​based on the eigenvalues ​​of the correlation coefficient matrix; Determine a corresponding target eigenvector based on the target eigenvalue to obtain an eigenvector matrix; The dimension reduction matrix is ​​obtained based on the normalized matrix and the eigenvector matrix.

5. The method according to claim 4, characterized in that The step of normalizing the original matrix to obtain a normalized matrix includes: Standardize each element of the original matrix based on the mean and standard deviation of each evaluation index in the original matrix to obtain a standardized matrix; The step of obtaining the dimension reduction matrix based on the normalized matrix and the eigenvector matrix includes: For each column of data in the standardized matrix, each element in each column of data is normalized based on the index mean of each column of data to obtain a sample matrix; The product of the sample matrix and the eigenvector matrix is ​​used as the dimension reduction matrix.

6. The method according to claim 1, characterized in that The grey correlation method is used to correlate the initial expert weights of the evaluation indicators in the dimension reduction matrix to obtain the indicator weights of the evaluation indicators, including: Based on the weight score of each expert for each evaluation indicator in the dimensionality reduction matrix, an initial expert weight matrix corresponding to the dimensionality reduction matrix is ​​obtained, and each column of the initial expert weight matrix is ​​used as a comparison data column; Screening out the maximum value of each column from the initial expert weight matrix to form a reference data column; Based on the Mahalanobis distance between each comparison data column and the reference data column, an indicator weight of the evaluation indicator corresponding to each comparison data column is determined.

7. The method according to claim 6, characterized in that The calculation formula of the indicator weight of the evaluation indicator is: Among them, ω k represents the index weight of the kth evaluation index, ω′ k Represents ω k The data before normalization, E0 represents the reference data column, E k represents the kth comparison data column, k=1, 2, ..., p, p is the number of evaluation indicators in the dimension reduction matrix; D s represents the Mahalanobis distance.

8. The method according to claim 5, characterized in that The clustering process of the values ​​of the evaluation indicators in the dimension reduction matrix includes: Performing maximum-minimum normalization processing on each element of the dimension reduction matrix to obtain a standardized dimension reduction matrix; The row vectors of the standardized dimensionality reduction matrix are used as data points, and the distance between two row vectors is used as the distance between data points. The row vectors in the standardized dimensionality reduction matrix are clustered using a density-based clustering algorithm to obtain the center point of each cluster cluster and its coordinates; the center point of the cluster cluster is used to indicate the average score of each energy storage power station in the cluster cluster, and the coordinate of the center point in each evaluation index dimension is used to indicate the average score of all energy storage power stations in the corresponding cluster cluster for each evaluation index.

9. The method according to claim 8, characterized in that The evaluation of each energy storage power station in each cluster based on the value of each evaluation index in each cluster in the clustering result and the corresponding index weight includes: For each cluster, the product of the coordinates of the center point of the cluster in each evaluation index dimension and the corresponding index weight is accumulated to obtain a comprehensive score for each cluster; The energy storage power stations within each cluster are evaluated based on the comprehensive score of each cluster.

10. A large-scale energy storage power station evaluation system based on multi-dimensional data fusion, characterized in that: include: A data fusion module is used to calculate the values ​​of each evaluation index corresponding to multiple evaluation angles based on the multi-dimensional operation data of each energy storage power station in the large-scale energy storage power station obtained, and fuse the values ​​of each evaluation index corresponding to the multiple evaluation angles into an original matrix of the evaluation index of the large-scale energy storage power station based on the energy storage power station; A dimension reduction module is used to perform dimension reduction processing on the original matrix using a principal component analysis method to obtain a corresponding dimension reduction matrix; A weight determination module is used to perform correlation processing on the expert weights of each evaluation index in the dimension reduction matrix using a grey correlation method to obtain an index weight of each evaluation index; The evaluation module is used to cluster the values ​​of the evaluation indicators in the dimension reduction matrix, and evaluate the energy storage power stations in each cluster based on the values ​​of the evaluation indicators in each cluster and the corresponding indicator weights in the cluster results.

Images