Evaluation system of power distribution network control system

Abstract

The invention relates to the technical field of power distribution networks, and discloses a power distribution network control system evaluation system, which comprises a data acquisition module used for acquiring power consumption data of a target area according to a fixed time interval in a first preset time period; the data clustering module is used for clustering the power utilization data of the target area in a first preset time period to obtain a plurality of characteristic power utilization time periods; the shielding module is used for carrying out at least one power distribution station shielding experiment on each characteristic power utilization time period to obtain evaluation data of each characteristic power utilization time period; the data processing module is used for performing multi-dimensional analysis on the evaluation data to obtain an evaluation result of the power distribution network control system in each dimension; and the comprehensive scoring module is used for obtaining a comprehensive score of the power distribution network control system according to the evaluation result of the power distribution network control system in each dimension.

Description

Technical Field

[0001] The present invention relates to the field of power distribution network, and more specifically, to a power distribution network control system evaluation system. Background Art

[0002] In modern distribution networks, the control system of distribution networks directly affects the stability of power supply. As the power grid structure becomes increasingly complex, the evaluation of traditional distribution network control systems is mainly based on a single indicator, such as power quality, which is difficult to fully reflect the comprehensive performance of complex distribution networks.

[0003] With the development of smart grids, distribution network control is gradually moving towards a multi-level and multi-dimensional complex system, which requires real-time multi-dimensional evaluation to fully grasp key performance such as load stability, reliability and power quality. Summary of the invention

[0004] The present invention provides a distribution network control system evaluation system to solve the technical problems raised in the background technology.

[0005] The present invention provides a distribution network control system evaluation system, comprising:

[0006] A data collection module, used to collect power consumption data of a target area at fixed time intervals within a first preset time period;

[0007] A data clustering module, used to cluster the power consumption data of the target area within a first preset time period to obtain a plurality of characteristic power consumption time periods;

[0008] A shielding module is used to perform at least one power distribution station shielding experiment for each characteristic power consumption time period to obtain evaluation data for each characteristic power consumption time period;

[0009] A data processing module is used to perform multi-dimensional analysis on the evaluation data to obtain multi-dimensional analysis results of the distribution network control system in each dimension;

[0010] The comprehensive scoring module is used to obtain a comprehensive score of the distribution network control system based on the evaluation results of the distribution network control system in each dimension.

[0011] Furthermore, the power consumption data of the target area within the first preset time period is clustered to obtain multiple characteristic power consumption time periods, including:

[0012] Dividing the power consumption data in the first preset time period into multiple periods to obtain power consumption data of multiple periods;

[0013] The minimum cycle is divided into fixed time intervals to obtain a number of time periods to be allocated;

[0014] Step 1: cluster several time periods to be allocated, as follows:

[0015] Step 2, among several time periods to be allocated, initialize k time periods to be allocated whose power consumption data differences are greater than a preset radius as cluster centers;

[0016] Step 3, calculating the difference between the a-th time period to be allocated and the b-th cluster center, and if it is determined that the difference is less than the preset radius, the a-th time period to be allocated is divided into the b-th cluster center;

[0017] Step 4, repeat step 3, divide several time periods to be allocated, and obtain k characteristic power consumption time periods;

[0018] Step 5: Calculate the mean of the power consumption data in each characteristic power consumption time period, and use the mean as the updated cluster center of the corresponding characteristic power consumption time period.

[0019] Furthermore, at least one distribution station shielding experiment is performed for each characteristic power consumption time period to obtain evaluation data for each characteristic power consumption time period, including:

[0020] Step 6, obtaining the updated cluster center of the target area in the sth characteristic power consumption time period;

[0021] Step 7, obtaining the number of distribution stations in the target area and the maximum output power of each distribution station;

[0022] Step 8: According to the number of distribution stations and the maximum output power of each distribution station, and the updated cluster center of the sth characteristic power consumption time period, determine the minimum number of distribution stations required in the target area in any time period within the sth characteristic power consumption time period, and reduce the distribution stations in the target area to the minimum number of distribution stations required;

[0023] Step 9: linearly reduce the output power of the distribution stations that are determined to be reduced to 0 simultaneously within the corresponding time period, and collect evaluation data of the remaining distribution stations in the target area.

[0024] Furthermore, the evaluation data was analyzed in multiple dimensions, including:

[0025] Step 10, according to the connection relationship of the distribution stations and the load fluctuation value of the distribution stations in the evaluation data;

[0026] Step 11: Construct the adjacency matrix A based on the connection relationship of the distribution stations load , the element in the i-th row and j-th column of the adjacency matrix represents the correlation between distribution stations i and j, and the correlation is expressed as 0 or 1;

[0027] Step 12: construct an initial load fluctuation vector M according to the load fluctuation value of the distribution station load , the sth element in the initial load fluctuation vector represents the initial load fluctuation value of the sth distribution station;

[0028] Step 13: Construct the load propagation matrix P according to the adjacency matrix and the initial load fluctuation vector. load , the load propagation matrix calculation formula is as follows:

[0029] Among them, P load (t) represents the load propagation matrix between distribution stations at the tth moment; t represents the tth moment of the corresponding shielding experiment, ∞ represents an infinite number, and m represents the index of the infinite number; (A load ·t) m Indicates A load ·t raised to the power of m; (·)! represents the factorial operation;

[0030] Step 14, calculate P by iteration load (t) until convergence is reached at time T, satisfying P load (T)≈P load (T-1), then the converged load propagation matrix P is obtained steady The element in the cth row and dth column of the converged load propagation matrix represents the impact value of the cth distribution station on the dth distribution station.

[0031] Furthermore, multi-dimensional analysis of the evaluation data also includes:

[0032] Step 15: evaluate the target area in the data. The number of power outages in the shielding experiment during the characteristic power consumption time period is x i1 , Restoration time for each power outage x i2 and total outage time x i3 ;

[0033] Step 16, The shielding experiment The data is normalized, and the normalization formula is: in, It means that after normalization The first characteristic power consumption period in the shielding experiment Item data, Represents the original pair The first characteristic power consumption period in the shielding experiment Item data, Indicates that k shielding experiments are performed in A collection on item data, Indicates that k shielding experiments are performed in The minimum value in the item data, Indicates that k shielding experiments are performed in The maximum value in the data item, The data items include the number of power outages xi1 , Restoration time for each power outage x i2 and total outage time x i3 ;

[0034] Step 17, set the number of clusters C, and initialize the fuzzy membership matrix U. The element in the e-th row and f-th column of the fuzzy membership matrix represents the The membership degree of the e-th item of data in the shielding experiment for the characteristic power consumption time period belongs to the f-th cluster;

[0035] U e,f ∈[0,1], and Among them, U e,f Indicates The e-th item of data in the shielding experiment for the characteristic power consumption time period belongs to the f-th cluster, and g represents the index of the cluster number C;

[0036] Step 18: Initialize the center vector v of C clusters f ;

[0037] Step 19: Use the fuzzy C-means algorithm to update the center of each cluster. The calculation formula is as follows:

[0038] Among them, v f,r represents the cluster center value of the center vector of the fth cluster on the rth index, and the cluster center value on the rth index represents the average number of power outages, the average power outage recovery time or the average total power outage time of the k shielding experiments; δ represents the fuzzy parameter used to control the fuzziness degree of membership, 2≥δ>1;

[0039] Step 20, based on The number of power outages in the shielding experiment during the characteristic power consumption time period is x i1 , Restoration time for each power outage x i2 and total outage time x i3 and the center vector v of C clusters f The distance between the two is updated, and the membership matrix U is updated. The formula is as follows:

[0040] Among them, U e,f (new) indicates the updated The membership degree of the e-th item of data in the shielding experiment for the characteristic power consumption time period belongs to the f-th cluster; Indicates The distance between the center of the fth cluster in the shielding experiment of the characteristic power consumption time period, d e,f =‖x e -v f ‖, x e Indicates The e-th item of data in the shielding experiment is performed for characteristic power consumption time periods. The e-th item of data includes the number of power outages x i1 , Restoration time for each power outage x i2 and total outage time x i3 ;

[0041] Step 21, repeat step 17, and determine that the iteration difference of the membership matrix is ​​less than the preset iteration threshold, then determine that the membership matrix and the cluster center converge, and obtain the final fuzzy membership matrix. The element in the e-th row and f-th column of the final fuzzy membership matrix represents the membership matrix of the cluster center. The membership degree of the e-th item of data in the shielding experiment for characteristic power consumption time periods belongs to the f-th cluster.

[0042] Furthermore, multi-dimensional analysis of the evaluation data also includes:

[0043] Step 22, based on the evaluation data, obtain The voltage fluctuation data of each distribution station, the harmonic content of each distribution station and the power factor data of each distribution station in the shielding experiment during the characteristic power consumption period are as follows;

[0044] Step 23, constructing a power quality data matrix Q based on the voltage fluctuation data of each distribution station, the harmonic content of each distribution station and the power factor data of each distribution station, b represents the characteristic vector constructed by the harmonic content and power factor data of each distribution station, and u represents the characteristic vector of the first The number of time periods for collecting voltage fluctuation data of each distribution station, harmonic content of each distribution station and power factor data of each distribution station in the shielding experiment during the characteristic power consumption time period;

[0045] Step 24, multi-scale entropy calculation is performed on the power quality data in the power quality data matrix Q of each distribution station, and the calculation formula is as follows:

[0046] Among them, E o,p,q represents the complex entropy value of the oth distribution station in the qth time period under the pth power quality index; p o,p,q,l represents the probability distribution of the power quality data of the o-th distribution station in the q-th time period under the p-th power quality index on the l-th time scale, where l is an integer multiple of u, and L represents the number of time scales;

[0047] Step 25, according to the result of multi-scale entropy calculation of the power quality data of the distribution station, calculate the overall quality of the power quality data of the distribution station, and the calculation formula is as follows:

[0048] in, Represents the overall quality of the power quality data of the oth distribution station.

[0049] Furthermore, the multidimensional analysis results of the distribution network control system are weighted to obtain the comprehensive score of the distribution network control system, where the multidimensional analysis results include the converged load propagation matrix P steady , the membership matrix U and the overall quality of the power quality data of the distribution station are as follows:

[0050] Step 26, pass Get the comprehensive score S of the distribution network control system all ; where β1, β2 and β3 represent the first weight coefficient, the second weight coefficient and the third weight coefficient respectively, ‖·‖ F Represents an operation to find the Frobenius norm.

[0051] The beneficial effects of the present invention are as follows: through a multi-dimensional comprehensive evaluation system, combined with a time-divided shielding experiment design, a comprehensive and accurate performance evaluation of the distribution network control system is carried out from three key dimensions: load stability, power reliability and power quality, providing scientific data support for the optimal scheduling and stability improvement of the power grid, and effectively making up for the shortcomings of the traditional single indicator evaluation method. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a module diagram of a distribution network control system evaluation system of the present invention; DETAILED DESCRIPTION

[0053] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0054] like Figure 1 As shown, a distribution network control system evaluation system includes:

[0055] A data collection module, used to collect power consumption data of a target area at fixed time intervals within a first preset time period;

[0056] A data clustering module, used to cluster the power consumption data of the target area within a first preset time period to obtain a plurality of characteristic power consumption time periods;

[0057] A shielding module is used to perform at least one power distribution station shielding experiment for each characteristic power consumption time period to obtain evaluation data for each characteristic power consumption time period;

[0058] A data processing module is used to perform multi-dimensional analysis on the evaluation data to obtain multi-dimensional analysis results of the distribution network control system in each dimension;

[0059] The comprehensive scoring module is used to obtain a comprehensive score of the distribution network control system based on the evaluation results of the distribution network control system in each dimension.

[0060] In one embodiment of the present invention, the power consumption data of the target area within the first preset time period is clustered to obtain multiple characteristic power consumption time periods, including:

[0061] Dividing the power consumption data in the first preset time period into multiple periods to obtain power consumption data of multiple periods;

[0062] The minimum cycle is divided into fixed time intervals to obtain a number of time periods to be allocated;

[0063] Step 1: cluster several time periods to be allocated, as follows:

[0064] Step 2, among several time periods to be allocated, initialize k time periods to be allocated whose power consumption data differences are greater than a preset radius as cluster centers;

[0065] Step 3, calculating the difference between the a-th time period to be allocated and the b-th cluster center, and if it is determined that the difference is less than the preset radius, the a-th time period to be allocated is divided into the b-th cluster center;

[0066] Step 4, repeat step 3, divide several time periods to be allocated, and obtain k characteristic power consumption time periods;

[0067] Step 5: Calculate the mean of the power consumption data in each characteristic power consumption time period, and use the mean as the updated cluster center of the corresponding characteristic power consumption time period.

[0068] Specifically, in the evaluation process of the distribution network control system, in order to identify the time periods with different power consumption characteristics and capture the performance of the system under different load conditions, it is first necessary to perform cluster analysis on the power consumption data of the target area.

[0069] Data division by period: First, the power consumption data in the first preset time period is divided by period (such as by hour, day or week frequency) to obtain power consumption data of multiple periods. The purpose of this step is to facilitate the capture of periodic fluctuations and provide a basis for clustering in subsequent time periods.

[0070] Time segmentation within the minimum cycle: The power consumption data of each cycle is further segmented according to fixed time intervals to obtain several time periods to be allocated (such as 15-minute intervals within each hour), which are used to refine the clustering results to ensure that the power consumption characteristics within small time intervals can be accurately described.

[0071] Initialize cluster centers: among the several time periods to be allocated, first select several time periods with large differences in electricity consumption data as initial cluster centers to ensure that the initial cluster centers are highly representative, thereby improving clustering accuracy.

[0072] Time period clustering process: For each time period to be allocated, the difference between the time period and the existing cluster center is calculated based on the power consumption data of the time period. If the difference is less than the preset radius, the time period is divided into the corresponding cluster center; otherwise, its independence is retained or processed separately. By iteratively adjusting the attribution of each time period, the clustering results are gradually formed.

[0073] Update cluster centers: After the initial clustering is completed, the mean of the power consumption data of all time periods in each cluster is calculated and used as the update center of the cluster to more accurately describe the power consumption characteristics of each characteristic power consumption time period. The purpose of this step is to enable each cluster center to reflect the overall characteristics of the time period within the cluster.

[0074] Use of clustering results: Through the above clustering analysis steps, we finally get several characteristic power consumption time periods, each of which represents a typical load pattern. Shielding experiments and data collection are performed on these time periods respectively, which can effectively capture the performance of the distribution network system under different load conditions and provide data support for subsequent system evaluation and optimization.

[0075] In one embodiment of the present invention, at least one power distribution station shielding experiment is performed for each characteristic power consumption time period to obtain evaluation data for each characteristic power consumption time period, including:

[0076] Step 6, obtaining the updated cluster center of the target area in the sth characteristic power consumption time period;

[0077] Step 7, obtaining the number of distribution stations in the target area and the maximum output power of each distribution station;

[0078] Step 8: According to the number of distribution stations and the maximum output power of each distribution station, and the updated cluster center of the sth characteristic power consumption time period, determine the minimum number of distribution stations required in the target area in any time period within the sth characteristic power consumption time period, and reduce the distribution stations in the target area to the minimum number of distribution stations required;

[0079] Step 9: linearly reduce the output power of the distribution stations that are determined to be reduced to 0 simultaneously within the corresponding time period, and collect evaluation data of the remaining distribution stations in the target area.

[0080] Specifically, obtain the power consumption characteristics of the target area in a specific cluster: After completing the clustering of the power consumption data of the target area, first obtain the center point of each cluster, that is, the typical load evaluation data of each characteristic power consumption time period (that is, the average power consumption level in this time period). This step ensures that the shielding experiment is based on different load states and can reflect the performance of the system under different load conditions.

[0081] Statistics of the number of distribution stations and maximum output power: In the shielding experiment design, it is necessary to count the number of distribution stations in the target area and the maximum output power of each distribution station in order to accurately control the power supply capacity in the shielding experiment. This data provides a basis for subsequent distribution station regulation and control, ensuring that the power demand response under load reduction or fault conditions can be accurately shielded in the shielding experiment.

[0082] Determine the minimum required number of distribution stations: Based on the number of distribution stations in the target area and the maximum output power of each distribution station, calculate the minimum number of distribution stations required in a specific load period, and reduce the number of distribution stations in actual operation to the minimum required value. This operation shields the operation of the power grid under low redundancy configuration and is used to test the stability and reliability of the system under certain load requirements.

[0083] Gradually reduce the output power and collect evaluation data: After reducing the number of distribution stations, gradually reduce the output power of the remaining distribution stations and collect evaluation data of each distribution station during the gradual load reduction process. These data include but are not limited to: load fluctuations, voltage changes, fault occurrence rate, etc. This step can deeply observe the dynamic response of the power grid under different load pressures and record the changes in various system parameters.

[0084] Evaluate system performance: By analyzing the data collected during the shielding experiment, the system performance under different load conditions and shielding configurations can be evaluated. In particular, observe the stability of the system when gradually reducing the load and output power, as well as the power supply guarantee to the target area. These shielding experiment data provide an important reference for subsequent system performance evaluation and optimization.

[0085] The practical significance of the shielding test results: Through the above-mentioned shielding test method, the reliability and stability of the system can be tested under various load states and operating conditions, providing a basis for subsequent control strategy adjustments. This shielding test method ensures the comprehensiveness and representativeness of the evaluation data, so that the output results of the evaluation system can more realistically and accurately reflect the actual performance of the distribution network system.

[0086] In one embodiment of the present invention, a multi-dimensional analysis is performed on the evaluation data, including:

[0087] Step 10, according to the connection relationship of the distribution stations and the load fluctuation value of the distribution stations in the evaluation data;

[0088] Step 11: Construct the adjacency matrix A based on the connection relationship of the distribution stations load , the element in the i-th row and j-th column of the adjacency matrix represents the correlation between distribution stations i and j, and the correlation is expressed as 0 or 1;

[0089] Step 12: construct an initial load fluctuation vector M according to the load fluctuation value of the distribution station load , the sth element in the initial load fluctuation vector represents the initial load fluctuation value of the sth distribution station;

[0090] Step 13: Construct the load propagation matrix P according to the adjacency matrix and the initial load fluctuation vector. load , the load propagation matrix calculation formula is as follows:

[0091] Among them, P load (t) represents the load propagation matrix between distribution stations at the tth moment; t represents the tth moment of the corresponding shielding experiment, ∞ represents an infinite number, and m represents the index of the infinite number; (A load ·t) m Indicates A load ·t raised to the power of m; (·)! represents the factorial operation;

[0092] Step 14, calculate P by iteration load (t) until convergence is reached at time T, satisfying P load (T)≈P load (T-1), then the converged load propagation matrix P is obtained steady The element in the cth row and dth column of the converged load propagation matrix represents the impact value of the cth distribution station on the dth distribution station.

[0093] Specifically, the connection relationship and load fluctuation value of the distribution station are obtained: in the multidimensional analysis, the connection relationship and load fluctuation value of each distribution station are first extracted from the shielding experiment data. The connection relationship indicates the physical or logical connection between the distribution stations, providing structural information for the construction of the load propagation matrix; the load fluctuation value indicates the load change of each distribution station under the shielding experiment conditions, which is used to reflect the response characteristics of the system under different load conditions.

[0094] Constructing the adjacency matrix: Generate an adjacency matrix based on the connection relationship between the distribution stations. Usually, if there is a direct connection between the two distribution stations, the matrix element is assigned a value of 1, otherwise it is 0. The establishment of the adjacency matrix provides the basis for the load propagation path.

[0095] Constructing the initial load fluctuation vector: Using the load fluctuation values ​​of each distribution station, construct the initial load fluctuation vector, where each element represents the initial load fluctuation of a specific distribution station. This vector is used to shield the propagation process of load fluctuation in the distribution network and is the basic data for the calculation of the load propagation matrix.

[0096] Based on the adjacency matrix and the load fluctuation vector, the load propagation matrix is ​​calculated. Each element of the matrix represents the propagation effect of the load fluctuation between distribution stations. The dynamic diffusion of the load fluctuation in the system is shielded in the form of matrix exponential (infinite series summation).

[0097] The load propagation matrix is ​​iteratively calculated until the system reaches a stable state, that is, the propagation matrix tends to converge after time T. The element values ​​in the converged load propagation matrix represent the degree of influence of load fluctuations on each distribution station in a stable state.

[0098] In one embodiment of the present invention, performing multidimensional analysis on the evaluation data further includes:

[0099] Step 15: evaluate the target area in the data. The number of power outages in the shielding experiment during the characteristic power consumption time period is x i1 , Restoration time for each power outage x i2 and total outage time x i3 ;

[0100] Step 16, The shielding experiment The data is normalized, and the normalization formula is: in, It means that after normalization The first characteristic power consumption period in the shielding experiment Item data, Represents the original pair The first characteristic power consumption period in the shielding experiment Item data, Indicates that k shielding experiments are performed in A collection on item data, Indicates that k shielding experiments are performed in The minimum value in the item data, Indicates that k shielding experiments are performed in The maximum value in the data item, The data items include the number of power outages x i1 , Restoration time for each power outage x i2 and total outage time x i3 ;

[0101] Step 17, set the number of clusters C, and initialize the fuzzy membership matrix U. The element in the e-th row and f-th column of the fuzzy membership matrix represents the The membership degree of the e-th item of data in the shielding experiment for the characteristic power consumption time period belongs to the f-th cluster;

[0102] U e,f ∈[0,1], and Among them, U e,f Indicates The e-th item of data in the shielding experiment for the characteristic power consumption time period belongs to the f-th cluster, and g represents the index of the cluster number C;

[0103] Step 18: Initialize the center vector v of C clusters f ;

[0104] Step 19: Use the fuzzy C-means algorithm to update the center of each cluster. The calculation formula is as follows:

[0105] Among them, v f,r represents the cluster center value of the center vector of the fth cluster on the rth index, and the cluster center value on the rth index represents the average number of power outages, the average power outage recovery time or the average total power outage time of the k shielding experiments; δ represents the fuzzy parameter used to control the fuzziness degree of membership, 2≥δ>1;

[0106] Step 20, based on The number of power outages in the shielding experiment during the characteristic power consumption time period is x i1 , Restoration time for each power outage x i2 and total outage time x i3 and the center vector v of C clusters f The distance between the two is updated, and the membership matrix U is updated. The formula is as follows:

[0107] Among them, U e,f (new) indicates the updated The membership degree of the e-th item of data in the shielding experiment for the characteristic power consumption time period belongs to the f-th cluster; Indicates The distance between the center of the fth cluster in the shielding experiment of the characteristic power consumption time period, d e,f =‖x e -v f ‖, x e Indicates The e-th item of data in the shielding experiment is performed for characteristic power consumption time periods. The e-th item of data includes the number of power outages x i1 , Restoration time for each power outage x i2 and total outage time xi3 ;

[0108] Step 21, repeat step 17, and determine that the iteration difference of the membership matrix is ​​less than the preset iteration threshold, then determine that the membership matrix and the cluster center converge, and obtain the final fuzzy membership matrix. The element in the e-th row and f-th column of the final fuzzy membership matrix represents the membership matrix of the cluster center. The membership degree of the e-th item of data in the shielding experiment for characteristic power consumption time periods belongs to the f-th cluster.

[0109] Specifically, the data such as power outage time, number of power outages, and restoration time collected during the shielding experiment are clustered and analyzed through fuzzy clustering technology to quantify the reliability performance of the system under different load conditions.

[0110] Obtaining power outage data in the target area: First, under different shielding test conditions, collect power outage-related data in the target area, including the total time of power outages, the number of power outages, and the recovery time of each power outage. These data can reflect the reliability of the system under different load or control conditions.

[0111] Determine the number of clusters: According to the shielding experiment design and the actual needs of the system, the reliability of the system is divided into several levels (such as high reliability, medium reliability and low reliability), corresponding to the number of clusters. The determination of the number of clusters helps the subsequent calculation of membership, so that each distribution station can be accurately assigned to the corresponding reliability level.

[0112] Initialize the fuzzy membership matrix: In the fuzzy cluster analysis, a fuzzy membership matrix is ​​initialized for each substation; the membership value ranges from 0 to 1, and the sum of the membership values ​​of each substation in all clusters is 1.

[0113] Calculate the initial cluster center: Use the fuzzy mean algorithm to determine each cluster center (i.e., the center value of the reliability level). Each center consists of the mean of the system's power outage time, power outage number, and recovery time. The calculation of the initial cluster center ensures the accuracy of the clustering process and helps to divide the system reliability level.

[0114] Update the membership matrix: For each distribution station, update the membership according to its distance from each cluster center. Iterative update of cluster centers: Recalculate the location of cluster centers according to the updated membership matrix. Through multiple iterations, the cluster centers and the membership matrix converge simultaneously, and a stable clustering result is obtained, that is, the reliability membership of each distribution station.

[0115] Interpretation of the membership matrix: The converged membership matrix represents the reliability performance of each substation in the system. A higher membership value indicates that the substation is more stable at that reliability level. For example, if the membership of a substation in the "high reliability" cluster is close to 1, it means that it has high reliability under all shielding test conditions.

[0116] In one embodiment of the present invention, performing multidimensional analysis on the evaluation data further includes:

[0117] Step 22, based on the evaluation data, obtain The voltage fluctuation data of each distribution station, the harmonic content of each distribution station and the power factor data of each distribution station in the shielding experiment during the characteristic power consumption period are as follows;

[0118] Step 23, constructing a power quality data matrix Q based on the voltage fluctuation data of each distribution station, the harmonic content of each distribution station and the power factor data of each distribution station, b represents the characteristic vector constructed by the harmonic content and power factor data of each distribution station, and u represents the characteristic vector of the first The number of time periods for collecting voltage fluctuation data of each distribution station, harmonic content of each distribution station and power factor data of each distribution station in the shielding experiment during the characteristic power consumption time period;

[0119] Step 24, multi-scale entropy calculation is performed on the power quality data in the power quality data matrix Q of each distribution station, and the calculation formula is as follows:

[0120] Among them, E o,p,q represents the complex entropy value of the oth distribution station in the qth time period under the pth power quality index; p o,p,q,l represents the probability distribution of the power quality data of the o-th distribution station in the q-th time period under the p-th power quality index on the l-th time scale, where l is an integer multiple of u, and L represents the number of time scales;

[0121] Step 25, according to the result of multi-scale entropy calculation of the power quality data of the distribution station, calculate the overall quality of the power quality data of the distribution station, and the calculation formula is as follows:

[0122] in, Represents the overall quality of the power quality data of the oth distribution station.

[0123] Specifically, by calculating the complexity entropy, the power quality performance of each distribution station in the system can be analyzed, including the dynamic changes of voltage fluctuations and harmonic content. The stability of the system power quality can be quantified and analyzed through the multi-scale entropy calculation method.

[0124] Power quality data collection: Under different shielding test conditions, the power quality data of each distribution station is collected, including:

[0125] Voltage fluctuation data: indicates the voltage changes of each distribution station during the shielding experiment, which can reflect the stability of the voltage.

[0126] Current harmonic content: Record the harmonic content of different frequency bands in the current to evaluate the cleanliness and stability of power quality.

[0127] Power factor: It indicates the effectiveness of power transmission and the efficiency of power transmission. It is an important indicator of power quality.

[0128] Construct a power quality data matrix: Based on the collected data, construct a three-dimensional power quality data matrix.

[0129] Multiscale entropy calculation: In order to accurately reflect the complexity and volatility of power quality, the multiscale entropy method is used to calculate the power quality data of each distribution station. Multiscale entropy can describe the complexity of power quality fluctuations on different time scales. A higher entropy value indicates that the power quality of the system fluctuates greatly, while a lower entropy value indicates that the power quality is relatively stable.

[0130] Construct the complexity entropy matrix: The multi-scale entropy calculation results of each distribution station are combined into a complexity entropy matrix, which is used to represent the power quality performance of the entire distribution network system under different time scales and shielding experimental conditions.

[0131] Evaluate the power quality stability of the system: By analyzing the entropy value in the complexity entropy matrix, the power quality performance of the distribution network system can be intuitively evaluated;

[0132] Higher complexity entropy: If the complexity entropy value of a distribution station is high, it means that its power quality fluctuates greatly and there may be unstable factors such as frequent voltage fluctuations or harmonic interference.

[0133] Lower complexity entropy: A lower complexity entropy value indicates that the power quality of the distribution station is relatively stable with smaller fluctuations.

[0134] Entropy mean calculation under shielding test conditions: The complexity entropy under each shielding test condition is averaged to obtain the overall power quality complexity score under a specific shielding or control strategy;

[0135] The mean can be used as an overall score for power quality complexity, with a lower mean indicating better power quality stability.

[0136] In one embodiment of the present invention, the multi-dimensional analysis results of the distribution network control system are weighted respectively to obtain a comprehensive score of the distribution network control system, wherein the multi-dimensional analysis results include the converged load propagation matrix P steady , the membership matrix U and the overall quality of the power quality data of the distribution station are as follows:

[0137] Step 26, pass Get the comprehensive score S of the distribution network control system all ; where β1, β2 and β3 represent the first weight coefficient, the second weight coefficient and the third weight coefficient respectively, ‖·‖ F Represents an operation to find the Frobenius norm.

[0138] Specifically, a comprehensive scoring method is used to comprehensively score the multi-dimensional evaluation data of the distribution network control system in order to form an overall evaluation result of the system. This scoring method combines the evaluation results of the three dimensions of load stability, power reliability and power quality, and integrates them into a unified comprehensive score, so as to comprehensively measure the overall performance of the distribution network control system.

[0139] Acquisition of multi-dimensional evaluation data: Conduct independent multi-dimensional evaluations of the system in three dimensions: load stability, power reliability, and power quality, and obtain evaluation data for each dimension.

[0140] Stability of the load propagation matrix: This evaluation data reflects the propagation characteristics and stability of the system under load fluctuation conditions. The more stable the load propagation, the stronger the system's ability to carry load fluctuations.

[0141] Reliability fuzzy cluster center: This result is obtained through fuzzy cluster analysis method, which represents the reliability performance of the system under different shielding test conditions. The higher the reliability, the better the stability of the system under different load conditions.

[0142] Complexity entropy of power quality: The complexity of power quality is quantified through multi-scale entropy calculation. The lower the complexity entropy, the more stable the power quality of the system under shielding test conditions.

[0143] Weight setting: According to different evaluation requirements, weights are assigned to the evaluation data of the three dimensions of load stability, power reliability and power quality.

[0144] Construct a comprehensive scoring formula: multiply the evaluation results of each dimension by the corresponding weight and sum them up to form the system's comprehensive score S.

[0145] Stability term of the load propagation matrix: The load stability performance is reflected by taking the inverse of the Frobenius norm of the load propagation matrix. The larger the value, the stronger the system's ability to adapt to load fluctuations.

[0146] The reliability fuzzy clustering item reflects the reliability level of the system. The closer the value is to 1, the higher the system reliability is.

[0147] Power quality complexity entropy term: The stability of power quality is expressed by the inverse of the sum of power quality complexity entropy. The larger the value, the more stable the power quality.

[0148] Analysis of comprehensive score results: The overall performance of the distribution network control system can be intuitively evaluated through the calculated comprehensive score S. The higher the comprehensive score, the better the system performs in terms of load stability, reliability and power quality.

[0149] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. A distribution network control system evaluation system, characterized in that: include: A data collection module, used to collect power consumption data of a target area at fixed time intervals within a first preset time period; A data clustering module, used to cluster the power consumption data of the target area within a first preset time period to obtain a plurality of characteristic power consumption time periods; A shielding module is used to perform at least one power distribution station shielding experiment for each characteristic power consumption time period to obtain evaluation data for each characteristic power consumption time period; A data processing module is used to perform multi-dimensional analysis on the evaluation data to obtain multi-dimensional analysis results of the distribution network control system in each dimension; The comprehensive scoring module is used to weight the multi-dimensional analysis results of the distribution network control system respectively to obtain a comprehensive score of the distribution network control system.

2. A distribution network control system evaluation system according to claim 1, characterized in that: The power consumption data of the target area within the first preset time period is clustered to obtain multiple characteristic power consumption time periods, including: Dividing the power consumption data in the first preset time period into multiple periods to obtain power consumption data of multiple periods; The minimum cycle is divided into fixed time intervals to obtain a number of time periods to be allocated; Step 1: cluster several time periods to be allocated, as follows: Step 2, among several time periods to be allocated, initialize k time periods to be allocated whose power consumption data differences are greater than a preset radius as cluster centers; Step 3, calculating the difference between the a-th time period to be allocated and the b-th cluster center, and if it is determined that the difference is less than the preset radius, the a-th time period to be allocated is divided into the b-th cluster center; Step 4, repeat step 3, divide several time periods to be allocated, and obtain k characteristic power consumption time periods; Step 5: Calculate the mean of the power consumption data in each characteristic power consumption time period, and use the mean as the updated cluster center of the corresponding characteristic power consumption time period.

3. A distribution network control system evaluation system according to claim 2, characterized in that: Conduct at least one distribution station shielding experiment for each characteristic power consumption time period to obtain evaluation data for each characteristic power consumption time period, including: Step 6, obtaining the updated cluster center of the target area in the sth characteristic power consumption time period; Step 7, obtaining the number of distribution stations in the target area and the maximum output power of each distribution station; Step 8: According to the number of distribution stations and the maximum output power of each distribution station, and the updated cluster center of the sth characteristic power consumption time period, determine the minimum number of distribution stations required in the target area in any time period within the sth characteristic power consumption time period, and reduce the distribution stations in the target area to the minimum number of distribution stations required; Step 9: linearly reduce the output power of the distribution stations that are determined to be reduced to 0 simultaneously within the corresponding time period, and collect evaluation data of the remaining distribution stations in the target area.

4. A distribution network control system evaluation system according to claim 3, characterized in that: Conduct multi-dimensional analysis of assessment data, including: Step 10, according to the connection relationship of the distribution stations and the load fluctuation value of the distribution stations in the evaluation data; Step 11: Construct the adjacency matrix A based on the connection relationship of the distribution stations load , the element in the i-th row and j-th column of the adjacency matrix represents the correlation between distribution stations i and j, and the correlation is expressed as 0 or 1; Step 12: construct an initial load fluctuation vector M according to the load fluctuation value of the distribution station load , the sth element in the initial load fluctuation vector represents the initial load fluctuation value of the sth distribution station; Step 13: Construct the load propagation matrix P according to the adjacency matrix and the initial load fluctuation vector. load , the load propagation matrix calculation formula is as follows: Among them, P load (t) represents the load propagation matrix between distribution stations at the tth moment; t represents the tth moment of the corresponding shielding experiment, ∞ represents an infinite number, and m represents the index of the infinite number; (A load ·t) m Indicates A load ·t raised to the power of m; (·)! represents the factorial operation; Step 14, calculate P by iteration load (t) until convergence is reached at time T, satisfying P load (T)≈P load (T-1), then the converged load propagation matrix P is obtained steady The element in the cth row and dth column of the converged load propagation matrix represents the impact value of the cth distribution station on the dth distribution station.

5. A distribution network control system evaluation system according to claim 4, characterized in that: Multi-dimensional analysis of assessment data also includes: Step 15: evaluate the target area in the data. The number of power outages in the shielding experiment during the characteristic power consumption time period is x i1 , Restoration time for each power outage x i2 and total outage time x i3 ; Step 16, The shielding experiment The data is normalized, and the normalization formula is: in, It means that after normalization The first characteristic power consumption period in the shielding experiment Item data, Represents the original pair The first characteristic power consumption period in the shielding experiment Item data, Indicates that k shielding experiments are performed in A collection on item data, Indicates that k shielding experiments are performed in The minimum value in the item data, Indicates that k shielding experiments are performed in The maximum value in the data item, The data items include the number of power outages x i1 、Restoration time for each power outage x i2 and total outage time x i3 ; Step 17, set the number of clusters C, and initialize the fuzzy membership matrix U. The element in the e-th row and f-th column of the fuzzy membership matrix represents the The membership degree of the e-th item of data in the shielding experiment for the characteristic power consumption time period belongs to the f-th cluster; U e,f ∈[0,1], and Among them, U e,f Indicates The e-th item of data in the shielding experiment for the characteristic power consumption time period belongs to the f-th cluster, and g represents the index of the cluster number C; Step 18: Initialize the center vector v of C clusters f ; Step 19: Use the fuzzy C-means algorithm to update the center of each cluster. The calculation formula is as follows: Among them, v f,r represents the cluster center value of the center vector of the fth cluster on the rth index, and the cluster center value on the rth index represents the average number of power outages, the average power outage recovery time or the average total power outage time of the k shielding experiments; δ represents the fuzzy parameter used to control the fuzziness degree of membership, 2≥δ>1; Step 20, based on The number of power outages in the shielding experiment during the characteristic power consumption time period is x i1 , Restoration time for each power outage x i2 and total outage time x i3 and the center vector v of C clusters f The distance between the two is updated, and the membership matrix U is updated. The formula is as follows: Among them, U e,f (new) indicates the updated The membership degree of the e-th item of data in the shielding experiment for the characteristic power consumption time period belongs to the f-th cluster; Expressing the The distance between the center of the fth cluster in the shielding experiment of the characteristic power consumption time period, d e,f =‖x e -v f ‖, x e Indicates The e-th item of data in the shielding experiment is performed for characteristic power consumption time periods. The e-th item of data includes the number of power outages x i1 , Restoration time for each power outage x i2 and total outage time x i3 ; Step 21, repeat step 17, and determine that the iteration difference of the membership matrix is ​​less than the preset iteration threshold, then determine that the membership matrix and the cluster center converge, and obtain the final fuzzy membership matrix. The element in the e-th row and f-th column of the final fuzzy membership matrix represents the membership matrix of the cluster center. The membership degree of the e-th item of data in the shielding experiment for characteristic power consumption time periods belongs to the f-th cluster.

6. A distribution network control system evaluation system according to claim 5, characterized in that: Multi-dimensional analysis of assessment data also includes: Step 22, based on the evaluation data, obtain The voltage fluctuation data of each distribution station, the harmonic content of each distribution station and the power factor data of each distribution station in the shielding experiment during the characteristic power consumption period are as follows; Step 23, constructing a power quality data matrix based on the voltage fluctuation data of each distribution station, the harmonic content of each distribution station and the power factor data of each distribution station b represents the characteristic vector constructed by the harmonic content and power factor data of each distribution station, and u represents the characteristic vector of the first The number of time periods for collecting voltage fluctuation data of each distribution station, harmonic content of each distribution station and power factor data of each distribution station in the shielding experiment during the characteristic power consumption time period; Step 24, multi-scale entropy calculation is performed on the power quality data in the power quality data matrix Q of each distribution station, and the calculation formula is as follows: Among them, E o,p,q represents the complex entropy value of the oth distribution station in the qth time period under the pth power quality index; p o,p,q,l represents the probability distribution of the power quality data of the o-th distribution station in the q-th time period under the p-th power quality index on the l-th time scale, where l is an integer multiple of u, and L represents the number of time scales; Step 25, according to the result of multi-scale entropy calculation of the power quality data of the distribution station, calculate the overall quality of the power quality data of the distribution station, and the calculation formula is as follows: in, Represents the overall quality of the power quality data of the oth distribution station.

7. A distribution network control system evaluation system according to claim 6, characterized in that: The multidimensional analysis results of the distribution network control system are weighted respectively to obtain the comprehensive score of the distribution network control system. The multidimensional analysis results include the converged load propagation matrix P steady , the membership matrix U and the overall quality of the power quality data of the distribution station are as follows: Step 26, pass Get the comprehensive score S of the distribution network control system all ; where β1, β2 and β3 represent the first weight coefficient, the second weight coefficient and the third weight coefficient respectively, ‖·‖ F Represents an operation to find the Frobenius norm.

Images