County power distribution network operation state identification method and system based on fuzzy neural network
By introducing fuzzy neural network technology into county distribution networks, improving k-means algorithm and entropy weight method, and establishing a multi-dimensional evaluation index system, the problem of single data preprocessing and index system in traditional methods is solved, and the county distribution is achieved Accurate identification and monitoring of the operating status of the power grid improves the safety and stability of the operation of the power grid.
Patent Information
- Application Number
- CN202410965982.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-18
- Publication Date
- 2024-11-29
AI Technical Summary
Traditional algorithms lack effective data preprocessing and multi-dimensional evaluation in the operation status identification of county distribution networks, making it difficult to accurately reflect the nonlinearity and time-varying of the power grid, resulting in a decrease in the accuracy of the identification results. The existing index system is single and cannot fully reflect the power grid. Running status.
By introducing fuzzy neural network technology, using semi-supervised learning to improve k-means algorithm for correlation analysis, establish a multi-dimensional operating state evaluation index system, use entropy weight method to determine index weights, build an identification model based on fuzzy neural network, and perform data pre-data The processing and indicator system correction can achieve accurate identification of the operating status of county distribution networks.
It improves data processing efficiency and identification accuracy, can better adapt to dynamic changes of the power grid, provide more reliable technical support, and provide guarantees for the safe and stable operation of the power grid.
Smart Images

Figure CN119046631A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of power system operating status monitoring and identification, and in particular to a county distribution network operating status identification method and system based on fuzzy neural network. Background Art
[0002] With the massive access of distributed energy and the popularization of new loads such as electric vehicle charging piles and energy storage devices, the structure and operation mode of county distribution networks have become more complex. These changes have increased the uncertainty and volatility of grid operation. The traditional algorithm operation status identification has not fully considered the nonlinearity, time-varying and uncertainty of the grid, resulting in a decrease in the accuracy of the identification results when facing complex grid operation status. Moreover, the existing indicator system evaluates the distribution system mostly from the perspective of economic and technical benefits, lacks a comprehensive and integrated evaluation of the specific conditions of the county distribution network operation stage, and it is difficult to effectively identify the operation status of the county distribution network.
[0003] The shortcomings of existing technologies are mainly reflected in the following aspects: first, the traditional method lacks effective preprocessing steps in data processing, resulting in the vacancy values and abnormal values in the data having an adverse effect on the identification results; second, the existing evaluation index system is relatively single and cannot fully reflect the multi-dimensional operation status of the distribution network; third, the existing identification model has limited ability to process fuzzy and uncertain information, and it is difficult to accurately identify the real-time status of the distribution network. The county distribution network operation status identification method based on fuzzy neural network invented by us, by introducing fuzzy logic and neural network technology, is expected to achieve significant improvements in data processing efficiency, identification accuracy and adaptation to dynamic changes of the power grid, thereby providing more reliable technical support for the safe and stable operation of the distribution network. Summary of the invention
[0004] In view of the above-mentioned existing problems, in order to accurately identify the typical operating status of the county distribution network, the present invention establishes a multi-dimensional calculation model, uses the improved k-means algorithm of semi-supervised learning to perform correlation analysis to screen out indicators with high correlation coefficients, uses the entropy weight method to determine the comprehensive weight of the operating status indicators, and finally identifies the operating status by establishing a fuzzy neural network model.
[0005] In order to solve the above technical problems, a method for identifying the operating status of county distribution network based on fuzzy neural network is proposed, including:
[0006] The collected county distribution network data are preprocessed to remove missing values and outliers, and the data set is standardized to establish a multi-dimensional operation status evaluation index system; the correlation analysis is performed through the improved k-means algorithm of semi-supervised learning, the operation indicator state range of the county distribution network is corrected, and the county distribution network operation status index system is established; according to the county distribution network operation status index system, the entropy weight method is used to determine the comprehensive weight of the operation status evaluation index; a county distribution network identification model based on fuzzy neural network is established, and the operation status of the county distribution network is determined through sample learning of fuzzy rules.
[0007] As a preferred solution of the method for identifying the operating status of a county distribution network based on a fuzzy neural network described in the present invention, the multi-dimensional operating status evaluation index system includes establishing evaluation indicators through multiple dimensions of spare capacity margin, emergency transfer capability, real-time controllability, fault risk rate, and power supply quality to identify the typical operating status of the county distribution network;
[0008] The reserve capacity margin includes a safety margin and a maximum load margin, and the maximum load margin is:
[0009]
[0010] Among them, P u Available capacity is the total capacity that the grid can provide without violating any safety and operation constraints. max is the maximum load recorded at any point in time;
[0011] The safety margin is:
[0012]
[0013] Among them, P t is the load that needs to be transferred under fault conditions;
[0014] The emergency transfer capability includes determining the transfer path and available transfer capacity, and calculating the remaining capacity of all transfer routes under the current load as:
[0015] C s =C e -C n
[0016] Determine the transfer capacity, which is limited by the minimum remaining capacity in the transfer path:
[0017]
[0018] T z =min(C s )
[0019] Among them, Ce is the rated capacity of the line, C n is the current transmission power of the line;
[0020] The real-time controllability is:
[0021]
[0022] The controllability of distributed resources is:
[0023]
[0024] in, It is a controllable distributed resource. is the total amount of distributed resources;
[0025] The load controllability is:
[0026]
[0027] Among them, η total In order to comprehensively consider the real-time controllability evaluation of distributed resource controllability and load controllability, t is the evaluation time point, α(s) is the weight of the resource-load ratio at different time points, and η DG is the ratio of controllable distributed resources at time point s, η L is the controllable load ratio at time point s, Δs is the size of the time window, and w i is the weight of time point i, d i is the original data value at time point i, It means to filter the data at important time points. a and b are adjustment parameters to control e -b·s The rate of change over time s, i is the variable index, Indicates e -b·s The steepness of is the adjustment parameter, is the controllable load. is the total load;
[0028] The failure risk rate includes the equipment failure risk rate and the line-level failure risk rate. The equipment failure risk rate is:
[0029]
[0030] The line-level failure risk rate is:
[0031]
[0032] Among them, η eq represents the average failure rate of the equipment, n represents the number of equipment failures, T represents the equipment operation time, n is the statistical period, and L is the probability of power line failure in each month;
[0033] The power supply quality includes comprehensive consideration of voltage deviation, voltage imbalance and frequency deviation:
[0034]
[0035] Among them, η Q is the comprehensive deviation of power supply quality, T is the evaluation period, η vd is the voltage deviation, η vu is the voltage unbalance deviation, A is the number of sampling points of the frequency deviation, λ is the attenuation coefficient of the frequency deviation, η fd,i is the frequency deviation of the i-th sampling point, σ(η fd ) is the standard deviation of the frequency deviation, Ω(η fd ) is the frequency deviation weight function of the i-th sampling point, V t is the instantaneous voltage, is the average voltage.
[0036] As a preferred solution of the method for identifying the operating status of a county distribution network based on a fuzzy neural network according to the present invention, the correlation analysis includes randomly selecting k samples R=R1, R2, ..., R i ,R k} as the initial cluster center and calculate the sample x i The distance from the mean vector to find the distance from the sample x i The closest cluster R k :Distance=argmin k ||x i -R i || 2 , x i Enter the current R k cluster, and whether the constraint is violated. If the constraint is violated, execute R i =R i ∪{x i}, if the constraint is not violated, recalculate the distance and achieve the goal:
[0037] min(||x i -R i || 2 +||x j -R j || 2 )
[0038] Among them, i and j are variable indexes.
[0039] Calculate the cluster centers of k categories:
[0040]
[0041] Determine whether the cluster center changes:
[0042]
[0043] When ΔR k When >0, the cluster center changes, and the mean vector is updated to make the cluster center original R k,new Become a new R k,old , input to recalculate the new R of k categories k,new .
[0044] Perform iterative calculations until the change in cluster center is less than 1, stop iteration, repeat iterations until all samples are classified into clusters, and output clustering results.
[0045] Among them, R k is the kth cluster center, N k is the number of samples in the kth cluster, σ R is the variance of the sample distribution, ΔR k is the updated value of the kth cluster center, R k,new is the new cluster center, R k,old is the old cluster center, is the variance of the update value, β is the integral variable, and γ is the learning rate that controls the update step size.
[0046] As a preferred solution of the method for identifying the operating status of a county distribution network based on a fuzzy neural network according to the present invention, the correction of the operating indicator status interval of the county distribution network includes calculating the correlation coefficient θ of the indicator under each major category according to the clustering result:
[0047]
[0048] Among them, θ is the intra-class correlation coefficient of the kth cluster, E is the number of indicators, η 1i is the index value of the i-th sample in a cluster, is the average value of the index η1 of all samples in a cluster, η 2e is the index value of the e-th sample in another cluster, is the average value of the indicator η2 of all samples in another cluster, η1 and η2 can be the reserve capacity margin, emergency transfer capability, real-time controllability, failure risk rate and power supply quality.
[0049] Select the index with the highest correlation coefficient in (0,1] under each category to correct the operation index state interval of the county distribution network and determine the new operation state index interval η new :
[0050]
[0051]
[0052]
[0053] The county distribution network operation status indicator system is constructed based on the revised operation status indicator range.
[0054] Among them, η 1i is the difference between the index value of the i-th sample in a cluster and the cluster average value, μ i is the standard deviation of the indicator in the i-th cluster, i is the variable index, η 1,min is the lower limit of the corrected index interval, η 1,max It is the upper limit of an indicator range after correction.
[0055] As a preferred solution of the method for identifying the operating status of a county distribution network based on a fuzzy neural network described in the present invention, the comprehensive weight includes determining the comprehensive weight of the indicators in the system by using the entropy weight method according to the construction of the operating status indicator system of the county distribution network;
[0056] The original data is standardized, and the weight of each sample in each indicator is calculated for the standardized data. According to the weight of each indicator, the information entropy weight of each indicator is calculated:
[0057]
[0058] Among them, w j is the weight of the jth indicator, e j is the information entropy of the jth indicator, m is the total number of indicators, e l is the information entropy of the lth indicator.
[0059] As a preferred solution of the method for identifying the operating status of a county distribution network based on a fuzzy neural network according to the present invention, the county distribution network identification model includes constructing a county distribution network identification model based on a fuzzy neural network structure:
[0060]
[0061] Among them, F(S,η new ,Θ) is the output result of the county distribution network operation status identification model, S is the sample data set, η new is the corrected operating status index interval, Θ is the fuzzy neural network structure parameter, and M represents the number of indicators in the sample data set S, that is, the number of indicator weights in the county distribution network operating status index system Λ. i is the comprehensive weight of the ith indicator determined by the entropy weight method, S i is the actual value of the i-th indicator in the sample data set, ∈i and σ i are the mean and variance of the ith indicator, Λ j is the weight of the jth indicator in the county distribution network operation status indicator system Λ, and r represents the corrected operation status indicator interval η new The number of indicators in , j is the variable index, η newj is the actual value of the jth indicator in the corrected operating status indicator interval, cen j and ku j are the center and width of the jth index respectively, and O represents the number of neurons in the fuzzy neural network, that is, the number of network structure parameters Θ. o is the weight of the oth neuron in the fuzzy neural network, sj o is the input of the oth neuron, cen o and ku o are the center and width of the oth neuron respectively, and ψ is the S-type activation function.
[0062] F(S,η new The output result of ,Θ) is used as the input variable and converted into the membership degree of fuzzy set through fuzzification process. For each input sample, the matching degree of the rules set according to the county distribution network operation status indicator system is calculated. The minimum-maximum reasoning is used to activate the rule with the highest matching degree. The conclusion part of all activated rules is averaged and the fuzzy output is converted into a specific output value to obtain the final result of the identification model. Each time a new sample is input, the sample's tolerable boundary and system error are calculated:
[0063]
[0064]
[0065] Among them, BJ(S) is the accommodated boundary, shape is the shape parameter, SP is the scale parameter, Γ is the gamma function, s is the integral variable, W is the system error function, G i is the actual output of the i-th sample.
[0066] When the accommodation boundary of the new sample is less than BJ(S)1 and the system error is less than W1, the model identification effect is considered to be good and no adjustment is required, and the operating state is directly divided according to the identification model.
[0067] When the accommodation boundary of the new sample is greater than BJ(S)1, the membership function in the fuzzy rule of the model is adjusted according to the degree of excess.
[0068] When the system error of the new sample is greater than W1, the county distribution network identification model of the neural network is optimized.
[0069] As a preferred scheme of the method for identifying the operating status of a county distribution network based on a fuzzy neural network described in the present invention, wherein: the determining the operating status of the county distribution network includes dividing the operating status of the county distribution network into normal operating status, warning status, emergency status, recovery status, maintenance status, fault status and reconstruction status.
[0070] The normal operating state means that all components of the power grid are operating within the new operating state index range of the county distribution network operating state index system without any faults or abnormalities. Under normal conditions, the power grid can meet the load demand, the power quality is good, and there is no voltage deviation or frequency deviation beyond the allowable limit.
[0071] The alert state is when the grid parameters are within the set allowable limit range or there are potential unstable factors, the grid switches to the alert state and takes preventive measures to prevent entering an emergency state, where the allowable limit range is within 0.6-0.8 times the new operating status index range of the county distribution network operating status index system.
[0072] The emergency state is when the parameters of the power grid exceed the prescribed allowable limit range by 10%, the power grid switches to the emergency state, immediately cuts off the load, isolates the faulty part, and avoids a large-scale system collapse.
[0073] The fault state is when a fault occurs in the power grid, resulting in interruption of power supply and degradation of quality. The power grid switches to the fault state, locates the fault point, and takes measures to isolate and repair it.
[0074] The recovery state is that after the emergency situation is lifted, the power grid transitions from the emergency state to the recovery state, and then to the normal operating state. After the fault state is lifted, the power grid switches from the fault state to the emergency state, then transitions to the recovery state, and continues to transition to the normal operating state. In the recovery state, the supply and demand are rebalanced, the damaged equipment is repaired, and the power grid structure is readjusted.
[0075] The maintenance state includes an isolation state. When maintaining and repairing equipment, the maintenance equipment of the power grid is in the maintenance state, and the associated equipment and associated areas of the maintenance equipment of the power grid are isolated and enter the isolation state.
[0076] The reconfiguration state is that the distribution network reconfigures the network structure by changing the switch state, optimizes the power distribution, and reduces the loss.
[0077] Another object of the present invention is to provide a county distribution network operation status identification system based on fuzzy neural network. The present invention first runs through the data preprocessing and index system construction module to provide standardized data and multi-dimensional evaluation index system. The improved k-means algorithm module of semi-supervised learning uses the preprocessed data to determine a more accurate index state interval through correlation analysis and clustering, and feeds the result back to the index system; the entropy weight method weight provides weight information for the fuzzy neural network, and the fuzzy neural network identification model module uses the data, index interval and weight provided by the first three modules to output the operation status of the county distribution network through training and learning, so as to accurately identify and monitor the operation status of the county distribution network, so as to adjust the power grid operation strategy in time and ensure the stable, safe and efficient operation of the power grid. Through continuous data processing and analysis, the system can provide early warning when abnormalities or risks occur in the distribution network, and assist decision makers to take appropriate measures.
[0078] As a preferred solution of the county distribution network operation status identification system based on fuzzy neural network described in the present invention, it is characterized by including a data preprocessing and indicator system construction module, a correction module and a power grid identification module.
[0079] The data preprocessing and indicator system construction module collects county distribution network data, performs preprocessing to remove missing values and abnormal values, standardizes the data set, establishes a multi-dimensional operation status evaluation indicator system, and provides basic data structure and evaluation standards.
[0080] The correction module performs correlation analysis through an improved k-means algorithm of semi-supervised learning, corrects the operating indicator status interval of the county distribution network, and establishes an operating status indicator system of the county distribution network.
[0081] The power grid identification module uses the entropy weight method to determine the comprehensive weight of the operating status evaluation index, provides a basis for weight distribution for the training and identification of the fuzzy neural network, establishes an identification model based on the fuzzy neural network, processes uncertainty information, learns through sample data, and outputs the operating status of the distribution network.
[0082] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the method described in the method for identifying the operating status of a county distribution network based on a fuzzy neural network are implemented.
[0083] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the method described in the method for identifying the operating status of a county distribution network based on a fuzzy neural network are implemented.
[0084] Beneficial effects of the present invention: The present invention establishes a multi-dimensional operating status evaluation index system, realizes data cleaning and standardization, ensures the accuracy and reliability of subsequent analysis, helps to improve the accuracy of county distribution network operating status identification, and lays a solid foundation for subsequent operating status evaluation; uses the entropy weight method to determine the comprehensive weight of the operating status evaluation index, realizes the objective assignment of the evaluation index weight, helps to improve the scientificity and rationality of the county distribution network operating status identification, makes the identification result more instructive, reduces the risk of human intervention, discovers potential problems, takes preventive measures in advance, and ensures the safe and stable operation of the power grid; the county distribution network identification model based on fuzzy neural network, through fuzzification process, minimum-maximum reasoning and fuzzy output conversion, intelligently identifies the operating status of the county distribution network, improves the identification speed and accuracy, and provides strong support for power grid dispatching. BRIEF DESCRIPTION OF THE DRAWINGS
[0085] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work, among which:
[0086] Figure 1 An overall flow chart of a method for identifying the operating status of a county distribution network based on a fuzzy neural network is provided as an embodiment of the present invention.
[0087] Figure 2 A system solution flow chart of a county distribution network operating status identification system based on fuzzy neural network provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0088] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.
[0089] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.
[0090] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is mutually exclusive with other embodiments, either individually or selectively.
[0091] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.
[0092] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.
[0093] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0094] Example 1, reference Figure 1 , which is the first embodiment of the present invention, provides a method for identifying the operating status of a county distribution network based on a fuzzy neural network, comprising:
[0095] S1: Preprocess the collected county distribution network data to remove missing values and outliers, standardize the data set, and establish a multi-dimensional operation status evaluation indicator system.
[0096] Furthermore, evaluation indicators are established through multiple dimensions such as spare capacity margin, emergency transfer capability, real-time controllability, fault risk rate, and power supply quality to identify the typical operating status of the county distribution network.
[0097] The reserve capacity margin includes a safety margin and a maximum load margin, and the maximum load margin is:
[0098]
[0099] Among them, P u Available capacity is the total capacity that the grid can provide without violating any safety and operation constraints. max is the maximum load recorded at any point in time;
[0100] The safety margin is:
[0101]
[0102] Among them, P t is the load that needs to be transferred under fault conditions;
[0103] The emergency transfer capability includes determining the transfer path and available transfer capacity, and calculating the remaining capacity of all transfer routes under the current load as:
[0104] C s =C e -C n
[0105] Determine the transfer capacity, which is limited by the minimum remaining capacity in the transfer path:
[0106]
[0107] T z =min(C s )
[0108] Among them, C e is the rated capacity of the line, C n is the current transmission power of the line;
[0109] The real-time controllability is:
[0110]
[0111] The controllability of distributed resources is:
[0112]
[0113] in, It is a controllable distributed resource. is the total amount of distributed resources;
[0114] The load controllability is:
[0115]
[0116] Among them, η total In order to comprehensively consider the real-time controllability evaluation of distributed resource controllability and load controllability, t is the evaluation time point, α(s) is the weight of the resource-load ratio at different time points, and η DGis the ratio of controllable distributed resources at time point s, η L is the controllable load ratio at time point s, Δs is the size of the time window, and w i is the weight of time point i, d i is the original data value at time point i, It means to filter the data of important time points. a and b are adjustment parameters to control e -b·s The rate of change over time s, i is the variable index, Indicates e -b·s The steepness of is the adjustment parameter, is the controllable load. is the total load;
[0117] The failure risk rate includes the equipment failure risk rate and the line-level failure risk rate. The equipment failure risk rate is:
[0118]
[0119] The line-level failure risk rate is:
[0120]
[0121] Among them, η eq represents the average failure rate of the equipment, n represents the number of equipment failures, T represents the equipment operation time, n is the statistical period, and L is the probability of power line failure in each month;
[0122] The power supply quality includes comprehensive consideration of voltage deviation, voltage imbalance and frequency deviation:
[0123]
[0124] Among them, η Q is the comprehensive deviation of power supply quality, T is the evaluation period, η vd is the voltage deviation, η vu is the voltage unbalance deviation, A is the number of sampling points of the frequency deviation, λ is the attenuation coefficient of the frequency deviation, η fd,i is the frequency deviation of the i-th sampling point, σ(η fd ) is the standard deviation of the frequency deviation, Ω(η fd ) is the frequency deviation weight function of the i-th sampling point, V t is the instantaneous voltage, is the average voltage.
[0125] S2: Through the improved k-means algorithm of semi-supervised learning, correlation analysis is performed to correct the operating indicator status range of the county distribution network and establish an operating status indicator system for the county distribution network.
[0126] Furthermore, k samples are randomly selected R = {R1, R2, ..., R i ,R k} as the initial cluster center and calculate the sample x i The distance from the mean vector to find the distance from the sample x i The closest cluster R k :Distance=argmin k ||x i -R i || 2 , x i Enter the current R k cluster, and whether the constraint is violated. If the constraint is violated, execute R i =R i ∪{x i}, if the constraint is not violated, recalculate the distance and achieve the goal:
[0127] min(||x i -R i || 2 +||x j -R j || 2 )
[0128] Among them, i and j are variable indexes.
[0129] Calculate the cluster centers of k categories:
[0130]
[0131] Determine whether the cluster center changes:
[0132]
[0133] When ΔR k When >0, the cluster center changes, and the mean vector is updated to make the cluster center original R k,new Become a new R k,old , input to recalculate the new R of k categories k,new .
[0134] Perform iteration calculations until the change in cluster center is less than 1, stop iteration, repeat iterations until all samples are classified into clusters, and output clustering results.
[0135] Among them, R k is the kth cluster center, N k is the number of samples in the kth cluster, σ R is the variance of the sample distribution, ΔR k is the updated value of the kth cluster center, R k,newis the new cluster center, R k,old is the old cluster center, is the variance of the update value, β is the integral variable, and γ is the learning rate that controls the update step size.
[0136] It should be noted that according to the clustering results, the correlation coefficient θ of the indicators under each major category is calculated:
[0137]
[0138] Among them, θ is the intra-class correlation coefficient of the kth cluster, E is the number of indicators, η 1i is the index value of the i-th sample in a cluster, is the average value of the index η1 of all samples in a cluster, η 2e is the index value of the e-th sample in another cluster, is the average value of the indicator η2 of all samples in another cluster, η1 and η2 can be the reserve capacity margin, emergency transfer capability, real-time controllability, failure risk rate and power supply quality.
[0139] Select the index with the highest correlation coefficient in (0,1] under each category to correct the operation index state interval of the county distribution network and determine the new operation state index interval η new :
[0140]
[0141]
[0142]
[0143] The county distribution network operation status indicator system is constructed based on the revised operation status indicator range.
[0144] Among them, η 1i is the difference between the index value of the i-th sample in a cluster and the cluster average value, μ i is the standard deviation of the indicator in the i-th cluster, i is the variable index, η 1,min is the lower limit of the corrected index interval, η 1,max It is the upper limit of an indicator range after correction.
[0145] S3: Based on the county distribution network operation status indicator system, the entropy weight method is used to determine the comprehensive weight of the operation status evaluation index.
[0146] Furthermore, according to constructing the county distribution network operation status indicator system, the comprehensive weight includes determining the comprehensive weight of the indicators in the system by using the entropy weight method according to constructing the county distribution network operation status indicator system;
[0147] The original data is standardized, and the weight of each sample in each indicator is calculated for the standardized data. According to the weight of each indicator, the information entropy weight of each indicator is calculated:
[0148]
[0149] Among them, w j is the weight of the jth indicator, e j is the information entropy of the jth indicator, m is the total number of indicators, e l is the information entropy of the lth indicator.
[0150] S4: Establish a county distribution network identification model based on fuzzy neural network, and determine the operating status of the county distribution network through sample learning fuzzy rules.
[0151] Furthermore, a county distribution network identification model based on fuzzy neural network structure is constructed:
[0152]
[0153] Among them, F(S,η new ,Θ) is the output result of the county distribution network operation status identification model, S is the sample data set, η new is the corrected operating status index interval, Θ is the fuzzy neural network structure parameter, and M represents the number of indicators in the sample data set S, that is, the number of indicator weights in the county distribution network operating status index system Λ. i is the comprehensive weight of the ith indicator determined by the entropy weight method, S i is the actual value of the i-th indicator in the sample data set, ∈ i and σ i are the mean and variance of the ith indicator, Λ j is the weight of the jth indicator in the county distribution network operation status indicator system Λ, and r represents the corrected operation status indicator interval η new The number of indicators in , j is the variable index, η newj is the actual value of the jth indicator in the corrected operating status indicator interval, cen j and ku j are the center and width of the jth index respectively, and O represents the number of neurons in the fuzzy neural network, that is, the number of network structure parameters Θ. o is the weight of the oth neuron in the fuzzy neural network, sj o is the input of the oth neuron, cen o and ku o are the center and width of the oth neuron respectively, and ψ is the S-type activation function.
[0154] F(S,ηnew The output result of ,Θ) is used as the input variable and converted into the membership degree of fuzzy set through fuzzification process. For each input sample, the matching degree of the rules set according to the county distribution network operation status indicator system is calculated. The minimum-maximum reasoning is used to activate the rule with the highest matching degree. The conclusion part of all activated rules is averaged and the fuzzy output is converted into a specific output value to obtain the final result of the identification model. Each time a new sample is input, the sample's tolerable boundary and system error are calculated:
[0155]
[0156] Among them, BJ(S) is the accommodated boundary, shape is the shape parameter, SP is the scale parameter, Γ is the gamma function, s is the integral variable, W is the system error function, G i is the actual output of the i-th sample.
[0157] When the accommodation boundary of the new sample is less than BJ(S)1 and the system error is less than W1, the model identification effect is considered to be good and no adjustment is required, and the operating state is directly divided according to the identification model.
[0158] When the accommodation boundary of the new sample is greater than BJ(S)1, the membership function in the fuzzy rule of the model is adjusted according to the degree of excess.
[0159] When the system error of the new sample is greater than W1, the county distribution network identification model of the neural network is optimized.
[0160] It should be noted that the operating status of the county distribution network is divided into normal operating status, warning status, emergency status, recovery status, maintenance status, fault status and reconstruction status.
[0161] The normal operating state means that all components of the power grid are operating within the new operating state index range of the county distribution network operating state index system without any faults or abnormalities. Under normal conditions, the power grid can meet the load demand, the power quality is good, and there is no voltage deviation or frequency deviation beyond the allowable limit.
[0162] The alert state is when the grid parameters are within the set allowable limit range or there are potential unstable factors, the grid switches to the alert state and takes preventive measures to prevent entering an emergency state, where the allowable limit range is within 0.6-0.8 times the new operating status index range of the county distribution network operating status index system.
[0163] The emergency state is when the parameters of the power grid exceed the prescribed allowable limit range by 10%, the power grid switches to the emergency state, immediately cuts off the load, isolates the faulty part, and avoids a large-scale system collapse.
[0164] The fault state is when a fault occurs in the power grid, resulting in interruption of power supply and degradation of quality. The power grid switches to the fault state, locates the fault point, and takes measures to isolate and repair it.
[0165] The recovery state is that after the emergency situation is lifted, the power grid transitions from the emergency state to the recovery state, and then to the normal operating state. After the fault state is lifted, the power grid switches from the fault state to the emergency state, then transitions to the recovery state, and continues to transition to the normal operating state. In the recovery state, the supply and demand are rebalanced, the damaged equipment is repaired, and the power grid structure is readjusted.
[0166] The maintenance state includes an isolation state. When maintaining and repairing equipment, the maintenance equipment of the power grid is in the maintenance state, and the associated equipment and associated areas of the maintenance equipment of the power grid are isolated and enter the isolation state.
[0167] The reconfiguration state is that the distribution network reconfigures the network structure by changing the switch state, optimizes the power distribution, and reduces the loss.
[0168] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.
[0169] Embodiment 2, the second embodiment of the present invention, provides a method for identifying the operating status of a county distribution network based on a fuzzy neural network. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.
[0170] The system collected a total of 10,000 data points. After removing missing values and outliers, 9,500 valid data points remained. An evaluation index system was established. By improving the k-means algorithm, the operating indicators were divided into five categories. When the change of the cluster center was less than 1, the iteration stopped; the weight of the standby capacity margin was 0.25, the weight of the emergency transfer capability was 0.2, the weight of the real-time controllability was 0.15, the weight of the fault risk rate was 0.2, and the weight of the power supply quality was 0.2.
[0171] The recognition model is established, and the fuzzy rules are determined through sample learning, and the model recognition effect is good.
[0172]
[0173] The spare capacity margin of the present invention is 0.7 of the prior art, and the emergency power transfer capability is improved by 41.67% compared with the prior art, indicating that the present invention has a higher response capability when dealing with sudden power grid conditions; the real-time controllability is better than 0.8 of the prior art, the failure risk rate is lower than 0.8 of the prior art, and the power supply quality is improved by 21.43% compared with the prior art, indicating that the present invention has obvious advantages in improving the power supply quality.
[0174] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.
[0175] Embodiment 3, the third embodiment of the present invention, is different from the first two embodiments in that:
[0176] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.
[0177] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.
[0178] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.
[0179] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0180] Example 4, reference Figure 2 , which is the fourth embodiment of the present invention, provides a county distribution network operation status identification system based on fuzzy neural network, including a data preprocessing and indicator system construction module, a correction module and a power grid identification module.
[0181] The data preprocessing and indicator system construction module collects county distribution network data, performs preprocessing to remove missing values and outliers, standardizes the data set, establishes a multi-dimensional operating status evaluation indicator system, and provides basic data structure and evaluation standards.
[0182] The correction module performs correlation analysis through the improved k-means algorithm of semi-supervised learning, corrects the operating indicator status range of the county distribution network, and establishes the operating status indicator system of the county distribution network.
[0183] The power grid identification module uses the entropy weight method to determine the comprehensive weight of the operating status evaluation index, provides a basis for weight allocation for the training and identification of the fuzzy neural network, establishes an identification model based on the fuzzy neural network, processes uncertainty information, learns through sample data, and outputs the operating status of the distribution network.
[0184] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.
Claims
1. A method for identifying the operating status of county distribution networks based on fuzzy neural networks. Features: include, The collected county distribution network data is preprocessed to remove missing values and outliers, and the data set is standardized to establish a multi-dimensional operating status evaluation indicator system; Through the improved k-means algorithm of semi-supervised learning, correlation analysis is performed to correct the operating indicator status range of the county distribution network and establish the operating status indicator system of the county distribution network. According to the county distribution network operation status index system, the comprehensive weight of the operation status evaluation index is determined by using the entropy weight method; A county distribution network identification model based on fuzzy neural network is established, and the operation status of the county distribution network is determined by learning fuzzy rules through samples.
2. The method for identifying the operating status of a county distribution network based on a fuzzy neural network as claimed in claim 1, Features: The multi-dimensional operation status evaluation index system includes establishing evaluation indicators through multiple dimensions of spare capacity margin, emergency transfer capability, real-time controllability, fault risk rate, and power supply quality to identify the typical operation status of the county distribution network; The reserve capacity margin includes a safety margin and a maximum load margin, and the maximum load margin is: Among them, P u Available capacity is the total capacity provided by the grid without violating any safety and operation constraints. max is the maximum load recorded at any point in time; The safety margin is: Among them, P t is the load that needs to be transferred under fault conditions; The emergency transfer capability includes determining the transfer path and available transfer capacity, and calculating the remaining capacity of all transfer routes under the current load as: C s =C e -C n Determine the transfer capacity, which is limited by the minimum remaining capacity in the transfer path: T z =min(C s ) Among them, C e is the rated capacity of the line, C n is the current transmission power of the line; The real-time controllability is: The controllability of distributed resources is: in, It is a controllable distributed resource. is the total amount of distributed resources; The load controllability is: Among them, η total In order to comprehensively consider the real-time controllability evaluation of distributed resource controllability and load controllability, t is the evaluation time point, α(s) is the weight of the resource-load ratio at different time points, and η DG is the ratio of controllable distributed resources at time point s, η L is the controllable load ratio at time point s, Δs is the size of the time window, and w i is the weight of time point i, d i is the original data value at time point i, It means to filter the data of important time points. a and b are adjustment parameters to control e -b·s The rate of change over time s, i is the variable index, Indicates e -b·s The steepness of is the adjustment parameter, is the controllable load. is the total load; The failure risk rate includes the equipment failure risk rate and the line-level failure risk rate. The equipment failure risk rate is: The line-level failure risk rate is: Among them, η eq represents the average failure rate of the equipment, n represents the number of equipment failures, T represents the equipment operation time, n is the statistical period, and L is the probability of power line failure in each month; The power supply quality includes comprehensive consideration of voltage deviation, voltage imbalance and frequency deviation: Among them, η Q is the comprehensive deviation of power supply quality, T is the evaluation period, η vd is the voltage deviation, η vu is the voltage unbalance deviation, A is the number of sampling points of frequency deviation, λ is the attenuation coefficient of frequency deviation, η fd,i is the frequency deviation of the i-th sampling point, σ(η fd ) is the standard deviation of the frequency deviation, Ω(η fd ) is the frequency deviation weight function of the i-th sampling point, V t is the instantaneous voltage, is the average voltage.
3. The method for identifying the operating status of a county distribution network based on a fuzzy neural network as claimed in claim 2, Features: The correlation analysis includes randomly selecting k samples R={R 1 ,R 2 ,…,R i ,R k } as the initial cluster center and calculate the sample x i The distance from the mean vector to find the distance from the sample x i The closest cluster R k :Distance=argmin k ||x i -R i || 2 , x i Enter the current R k cluster, and whether the constraint is violated. If the constraint is violated, execute R i =R i ∪{x i }, if the constraint is not violated, recalculate the distance and achieve the goal: min(||x i -R i || 2 +||x j -R j || 2 ) Among them, i and j are variable indexes; Calculate the cluster centers of k categories: Determine whether the cluster center changes: When ΔR k When >0, the cluster center changes, and the mean vector is updated to make the cluster center original R k,new Become a new R k,old , input to recalculate the new R of k categories k,new ; Perform iterative calculation until the change of cluster center is less than 1, stop iteration, repeat iteration until all samples are classified into clusters, and output clustering results; Among them, R k is the kth cluster center, N k is the number of samples in the kth cluster, σ R is the variance of the sample distribution, ΔR k is the updated value of the kth cluster center, R k,new is the new cluster center, R k,old is the old cluster center, is the variance of the update value, β is the integral variable, and γ is the learning rate that controls the update step size.
4. The method for identifying the operating status of a county distribution network based on a fuzzy neural network as claimed in claim 3, Features: The correction of the operating indicator state interval of the county distribution network includes calculating the correlation coefficient θ of the indicator under each major category according to the clustering result: Among them, θ is the intra-class correlation coefficient of the kth cluster, E is the number of indicators, η 1i is the index value of the i-th sample in a cluster, is the index η of all samples in a cluster 1 The average value, η 2e is the index value of the e-th sample in another cluster, is the index η of all samples in another cluster 2 The average value, η 1 and η 2 It can be used for spare capacity margin, emergency transfer capability, real-time controllability, failure risk rate and power supply quality; Select the index with the highest correlation coefficient in (0,1] under each category to correct the operation index state interval of the county distribution network and determine the new operation state index interval η new : According to the revised operating status index range, the county distribution network operating status index system is constructed; Among them, η 1i is the difference between the index value of the i-th sample in a cluster and the cluster average value, μ i is the standard deviation of the indicator in the i-th cluster, i is the variable index, η 1,min is the lower limit of the corrected index interval, η 1,max It is the upper limit of an indicator range after correction.
5. The method for identifying the operating status of a county distribution network based on a fuzzy neural network as claimed in claim 4, Features: The comprehensive weight includes constructing a county distribution network operation status indicator system and determining the comprehensive weight of the indicators in the system using the entropy weight method; The original data is standardized, and the weight of each sample in each indicator is calculated for the standardized data. According to the weight of each indicator, the information entropy weight of each indicator is calculated: Among them, w j is the weight of the jth indicator, e j is the information entropy of the jth indicator, m is the total number of indicators, e l is the information entropy of the lth indicator.
6. The method for identifying the operating status of a county distribution network based on a fuzzy neural network as claimed in claim 5, Features: The county distribution network identification model includes constructing a county distribution network identification model based on a fuzzy neural network structure: Among them, F(S,η new ,Θ) is the output result of the county distribution network operation status identification model, S is the sample data set, η new is the corrected operating status index interval, Θ is the fuzzy neural network structure parameter, M represents the number of indicators in the sample data set S, that is, the number of indicator weights in the county distribution network operating status index system Λ; w j is the comprehensive weight of the ith indicator determined by the entropy weight method, S i is the actual value of the i-th indicator in the sample data set, ∈ i and σ i are the mean and variance of the ith indicator, Λ j is the weight of the jth indicator in the county distribution network operation status indicator system Λ, and r represents the corrected operation status indicator interval η new The number of indicators in , j is the variable index, is the actual value of the jth indicator in the corrected operating status indicator interval, cen j and ku j are the center and width of the jth index respectively, O represents the number of neurons in the fuzzy neural network, that is, the number of network structure parameters Θ; o is the weight of the oth neuron in the fuzzy neural network, sj o is the input of the oth neuron, cen o and ku o are the center and width of the oth neuron, respectively, and ψ is the S-type activation function; F(S,η new The output result of ,Θ) is used as the input variable and converted into the membership degree of fuzzy set through fuzzification process. For each input sample, the matching degree of the rules set according to the county distribution network operation status indicator system is calculated. The minimum-maximum reasoning is used to activate the rule with the highest matching degree. The conclusion part of all activated rules is averaged and the fuzzy output is converted into a specific output value to obtain the final result of the identification model. Each time a new sample is input, the sample's tolerable boundary and system error are calculated: Among them, BJ(S) is the accommodated boundary, shape is the shape parameter, SP is the scale parameter, Γ is the gamma function, s is the integral variable, W is the system error function, G i is the actual output of the i-th sample; When the accommodated boundary of the new sample is less than BJ(S)1 and the system error is less than W1, the model identification effect is considered good and no adjustment is required, and the operating state is directly divided according to the identification model; When the accommodation boundary of the new sample is greater than BJ(S)1, the membership function in the fuzzy rule of the model is adjusted according to the degree of excess; When the system error of the new sample is greater than W1, the county distribution network identification model of the neural network is optimized.
7. The method for identifying the operating status of a county distribution network based on a fuzzy neural network as claimed in claim 6, Features: Determining the operating state of the county distribution network includes dividing the operating state of the county distribution network into normal operating state, alert state, emergency state, recovery state, maintenance state, fault state and reconstruction state; The normal operating state means that all components of the power grid are operating within the new operating state index range of the county distribution network operating state index system, without any fault or abnormality. Under normal conditions, the power grid can meet the load demand, the power quality is good, and there is no voltage deviation or frequency deviation beyond the allowable limit. The alert state is when the grid parameters are within the set allowable limit range or there are potential unstable factors, the grid switches to the alert state and takes preventive measures to prevent entering an emergency state, wherein the allowable limit range is within the range of 0.6-0.8 times the new operating status index range of the county distribution network operating status index system; The emergency state is when the parameters of the power grid exceed the specified allowable limit range by 10%, the power grid switches to the emergency state, immediately cuts off the load, isolates the faulty part, and avoids a large-scale system collapse; The fault state is when a fault occurs in the power grid, resulting in interruption of power supply and degradation of quality, the power grid switches to a fault state, locates the fault point and takes measures to isolate and repair it; The recovery state is that after the emergency is lifted, the power grid transitions from the emergency state to the recovery state, and then to the normal operation state. After the fault state is lifted, the power grid switches from the fault state to the emergency state, then transitions to the recovery state, and then continues to transition to the normal operation state. In the recovery state, the supply and demand are rebalanced, the damaged equipment is repaired, and the power grid structure is readjusted; The maintenance state includes an isolation state. When the equipment is maintained and repaired, the maintenance equipment of the power grid is in the maintenance state, and the associated equipment and associated areas of the maintenance equipment of the power grid are isolated and enter the isolation state; The reconfiguration state is that the distribution network reconfigures the network structure by changing the switch state, optimizes the power distribution, and reduces the loss.
8. A system using the method for identifying the operating status of a county distribution network based on a fuzzy neural network as claimed in any one of claims 1 to 7, Features: It includes data preprocessing and indicator system building module, correction module and power grid identification module; The data preprocessing and indicator system construction module collects county distribution network data, performs preprocessing to remove missing values and abnormal values, standardizes the data set, establishes a multi-dimensional operation status evaluation indicator system, and provides basic data structure and evaluation standards; The correction module performs correlation analysis through an improved k-means algorithm of semi-supervised learning, corrects the operating indicator status interval of the county distribution network, and establishes an operating status indicator system of the county distribution network; The power grid identification module uses the entropy weight method to determine the comprehensive weight of the operating status evaluation index, provides a basis for weight distribution for the training and identification of the fuzzy neural network, establishes an identification model based on the fuzzy neural network, processes uncertainty information, learns through sample data, and outputs the operating status of the distribution network.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program. It is characterized in that When the processor executes the computer program, the steps of the method for identifying the operating status of a county distribution network based on a fuzzy neural network according to any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the steps of the method for identifying the operating status of a county distribution network based on a fuzzy neural network according to any one of claims 1 to 7 are implemented.
Citation Information
Cited By
Energy storage power station state evaluation method and system based on real-time demand analysis
CN119671386A
County power distribution network multi-dimensional planning index flexible construction and evaluation method, system, device and medium
CN120387742A
A county power distribution network multi-dimensional planning index flexible construction and evaluation method, system, device and medium
CN120387742B
