Fault location method and system for power distribution network containing distributed photovoltaic and storage medium

Abstract

The application discloses a fault positioning method and system of a power distribution network containing distributed photovoltaic power and a storage medium, relates to the technical field of power distribution network fault positioning, and comprises the following steps: carrying out digital modeling and division on a target power distribution network containing distributed photovoltaic power to obtain a line unit sequence; when it is detected that the electrical parameters of any key switch are greater than or equal to a threshold value, obtaining the upstream and downstream abnormal switch electrical parameters of the upstream and downstream adjacent key switches, and calculating and obtaining the downstream abnormal switch electrical parameters after compensation; dividing the line unit sequence to obtain a plurality of hidden abnormal line units, carrying out line unit fault rate prediction to obtain a plurality of line unit fault rates; carrying out fault electrical parameter prediction to obtain a plurality of predicted upstream and downstream abnormal switch electrical parameters, respectively calculating the similarity, combining the plurality of line unit fault rates, and optimally selecting a fault line unit as a fault positioning result. The application solves the technical problem of inaccurate power distribution network fault positioning in the prior art.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power distribution network fault location, in particular to a fault location method and system for a power distribution network containing distributed photovoltaic power and a storage medium. BACKGROUND

[0002] In modern power systems, the reliability and power supply quality of the power distribution network directly affect the daily life of the general public and the normal operation of industrial production. With the transformation of energy structure and the rapid development of renewable energy, the penetration rate of distributed photovoltaic power in the power distribution network is continuously increasing, but its access also brings new problems to the fault location of the power distribution network.

[0003] In the prior art, the technical problem of low accuracy of fault location of the power distribution network containing distributed photovoltaic power needs to be solved. SUMMARY

[0004] The present application provides a fault location method and system for a power distribution network containing distributed photovoltaic power and a storage medium, which is used to solve the technical problem of inaccurate fault location of the power distribution network caused by the influence of the distributed photovoltaic power in the prior art.

[0005] In view of the above problems, the present application provides a fault location method and system for a power distribution network containing distributed photovoltaic power and a storage medium.

[0006] In a first aspect, the present application provides a fault location method for a power distribution network containing distributed photovoltaic power, which comprises:

[0007] The target power distribution network containing distributed photovoltaic power is digitally modeled, and line unit division is performed to obtain a line unit sequence, wherein the target power distribution network includes a plurality of key switches configured at a plurality of positions.

[0008] When it is detected that the switch electrical parameter at any one of the key switches is greater than or equal to an electrical parameter threshold value, the upstream abnormal switch electrical parameter and the downstream abnormal switch electrical parameter of the adjacent key switch downstream are obtained, and the compensated downstream abnormal switch electrical parameter is calculated and obtained.

[0009] A plurality of implicit abnormal line units are obtained by division in the line unit sequence, and the line unit fault rate is predicted according to the upstream abnormal switch electrical parameter and the compensated downstream abnormal switch electrical parameter to obtain a plurality of line unit fault rates.

[0010] According to the plurality of implicit abnormal line units, the fault electrical parameter is predicted to obtain a plurality of predicted upstream abnormal switch electrical parameters and a plurality of predicted downstream abnormal switch electrical parameters, the similarity with the upstream abnormal switch electrical parameter and the compensated downstream abnormal switch electrical parameter is calculated respectively, and the fault line unit is selected and obtained by optimization in combination with the plurality of line unit fault rates, as the fault location result.

[0011] In a second aspect, the application provides a fault location system for a power distribution network containing distributed photovoltaic, comprising:

[0012] a digital modeling module configured to perform digital modeling on a target power distribution network containing distributed photovoltaic, and to divide the target power distribution network into line units to obtain a sequence of line units, wherein the target power distribution network includes a plurality of key switches arranged at a plurality of positions.

[0013] a compensation parameter calculation module configured to, when detecting that the electrical parameter of any one of the key switches is greater than or equal to an electrical parameter threshold, obtain upstream abnormal switch electrical parameters and downstream abnormal switch electrical parameters of an adjacent key switch downstream of the upstream abnormal switch, and calculate compensated downstream abnormal switch electrical parameters.

[0014] a fault rate prediction module configured to divide the sequence of line units to obtain a plurality of implicitly abnormal line units, and to predict fault rates of the line units according to the upstream abnormal switch electrical parameters and the compensated downstream abnormal switch electrical parameters to obtain a plurality of line unit fault rates.

[0015] a fault location module configured to predict fault electrical parameters according to the plurality of implicitly abnormal line units to obtain a plurality of predicted upstream abnormal switch electrical parameters and a plurality of predicted downstream abnormal switch electrical parameters, to calculate similarities between the upstream abnormal switch electrical parameters and the compensated downstream abnormal switch electrical parameters, respectively, and to select a fault line unit as a fault location result by combining the plurality of line unit fault rates.

[0016] In a third aspect, the application provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the fault location method for a power distribution network containing distributed photovoltaic according to the first aspect.

[0017] The one or more technical solutions provided in the application have at least the following technical effects or advantages:

[0018] The application provides a fault location method, system and storage medium for a power distribution network containing distributed photovoltaic, effectively removes the pollution of photovoltaic random injected current on fault characteristic signals, restores the true electrical characteristics, probabilistically predicts potential fault line units, combines similarity verification, optimizes the decision by fusing fault probability and feature matching degree, and significantly improves the accuracy and robustness of fault location. Compared with the traditional method, the technical solution provided by the application significantly overcomes the change of current characteristics caused by the grid connection of distributed photovoltaic, and avoids false judgment of fault location.

[0019] The application achieves the technical effect of accurately and reliably identifying the fault section of the power distribution network containing distributed photovoltaic, guiding the operation and maintenance personnel to improve the fault handling efficiency and the safety of power grid operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in 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 ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 A schematic flow chart of a fault location method for a distributed photovoltaic distribution network provided in an embodiment of the present application.

[0022] Figure 2 A schematic diagram of the structure of a fault location system for a distributed photovoltaic distribution network provided in an embodiment of the present application.

[0023] Figure 3 A schematic diagram of the structure of the storage medium provided in an embodiment of the present application.

[0024] In the accompanying drawings, the components represented by the reference numerals are described as follows:

[0025] Digital modeling module 100 , compensation parameter calculation module 200 , failure rate prediction module 300 , fault location module 400 , computer readable storage medium 500 , and computer program 511 . DETAILED DESCRIPTION

[0026] The present application provides a fault location method, system and storage medium for a distributed photovoltaic distribution network, which is used to solve the technical problem in the prior art of inaccurate distribution network fault location caused by the influence of the distributed photovoltaic distribution network.

[0027] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0028] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0029] Example 1, as Figure 1 As shown, the present application provides a fault location method for a distributed photovoltaic distribution network, wherein the method includes:

[0030] S10: performing digital modeling on a target distribution network containing distributed photovoltaic, and performing line unit division to obtain a line unit sequence, wherein the target distribution network includes a plurality of key switches arranged at a plurality of positions.

[0031] In a distribution network with distributed photovoltaic access, the traditional fault location method usually adopts a simplified model, or does not fully consider the influence of the location of the distributed power supply on the current path, resulting in fuzzy potential fault area definition.

[0032] The step S10 in the method provided by the embodiments of the present application comprises:

[0033] Obtaining power grid characteristic information of the target distribution network, performing digital modeling to obtain a distribution network model, wherein the target distribution network contains distributed photovoltaic power supply and includes a plurality of key switches arranged at a plurality of positions.

[0034] Performing line unit division on the lines in the distribution network model to obtain a line unit sequence.

[0035] In the embodiments of the present application, the power grid characteristic information of the target distribution network, such as the position information and type information of the distribution equipment, is obtained, and a software such as DIgSILENT is used for digital modeling to obtain a distribution network model. The target distribution network contains distributed photovoltaic power supply and includes a plurality of key switches arranged at a plurality of positions.

[0036] The lines in the distribution network model are divided into line units to obtain a line unit sequence. The lines in the distribution network model are divided into several parts according to the branchless section to obtain line units, for example, each 50 meters is a line unit, and the line units are integrated and numbered in sequence to obtain a line unit sequence, which provides a basic unit for subsequent fault location.

[0037] By constructing an accurate digital model containing distributed photovoltaic and dividing the line unit sequence based on the topological relationship, a spatial analysis basis for fault location is established. This operation clearly defines the connection boundaries between the key switches in the distribution network, and the structured modeling effectively overcomes the problem of fuzzy fault section caused by rough network division in traditional methods, and provides a reliable physical framework support for subsequent accurate capture of abnormal signals.

[0038] S20: When it is detected that the switch electrical parameter at any one of the key switches is greater than or equal to the electrical parameter threshold, obtaining the upstream abnormal switch electrical parameter and the downstream abnormal switch electrical parameter of the adjacent key switch downstream, and calculating to obtain the compensated downstream abnormal switch electrical parameter.

[0039] The existing method directly uses the monitoring value to judge the fault when detecting the abnormality of the switch node electrical parameter, but the distributed photovoltaic may continue to supply power during the fault, polluting the electrical parameters of the downstream monitoring point, causing the downstream electrical quantity to fail to truly reflect the current change caused by the fault, resulting in distortion of the fault characteristics and a significant reduction in the reliability of positioning.

[0040] The step S20 in the method provided by the embodiment of the present application comprises:

[0041] The switch electrical parameters at the plurality of key switches are monitored, and when any one of the switch electrical parameters is greater than or equal to an electrical parameter threshold value, an upstream key switch is obtained, and the switch electrical parameter is taken as an upstream abnormal switch electrical parameter.

[0042] The switch electrical parameter of the downstream adjacent key switch in the current direction is obtained as a downstream abnormal switch electrical parameter.

[0043] According to the distributed photovoltaic operation and maintenance data, the distributed photovoltaic electrical parameter corresponding to the downstream adjacent key switch is indexed and obtained.

[0044] The downstream abnormal switch electrical parameter is obtained by subtracting the distributed photovoltaic electrical parameter from the downstream abnormal switch electrical parameter.

[0045] In the embodiment of the present application, according to the monitoring and collecting system of the power distribution network, the switch electrical parameters such as current parameters at the plurality of key switches are monitored. When any one of the switch electrical parameters is greater than or equal to an electrical parameter threshold value, an upstream key switch is obtained, that is, the line where the switch with the electrical parameter greater than or equal to the electrical parameter threshold value is searched, the upstream key switch closest to the switch with the electrical parameter greater than or equal to the electrical parameter threshold value is found as the closest key switch, and the electrical parameter of the upstream key switch is taken as the upstream abnormal switch electrical parameter, such as the current parameter of the upstream key switch. The electrical parameter threshold value is a threshold value reflecting the size of the electrical parameter set according to the rated electrical parameter. The electrical parameter threshold value is exemplarily set to 1.2 times the rated current. For example, if the rated current is 100 A, the electrical parameter threshold value = 100 x 1.2 = 1220 A.

[0046] The switch electrical parameter of the downstream adjacent key switch in the current direction is obtained as a downstream abnormal switch electrical parameter.

[0047] According to the distributed photovoltaic operation and maintenance data, the distributed photovoltaic electrical parameter corresponding to the downstream adjacent key switch is indexed and obtained.

[0048] When there is distributed photovoltaic, the downstream switch may have photovoltaic panels connected to the grid to continue power supply, resulting in inaccurate electrical parameters. The downstream abnormal switch electrical parameter is compensated by subtracting the distributed photovoltaic electrical parameter. When the downstream abnormal switch electrical parameter is 160A and the distributed photovoltaic electrical parameter is 80A, the compensated downstream abnormal switch electrical parameter is 160-80=80A.

[0049] The present application introduces an electrical parameter compensation mechanism for distributed photovoltaic. When the key switch triggers an anomaly, the real-time data of the photovoltaic associated by the index is used to dynamically deduct the photovoltaic injection component from the downstream abnormal switch electrical parameter to generate the compensated downstream abnormal switch electrical parameter. The pollution of the fault current characteristics caused by the continuous power supply of the distributed photovoltaic is effectively eliminated, the real electrical parameter change caused by pure fault is restored, and the distortion problem of the downstream monitoring data is solved, providing high-fidelity input data basis for subsequent fault probability prediction and feature matching.

[0050] S30: dividing to obtain a plurality of implied abnormal line units in the sequence of line units, and predicting the fault rate of the line units according to the upstream abnormal switch electrical parameter and the compensated downstream abnormal switch electrical parameter to obtain a plurality of line unit fault rates.

[0051] The traditional method relies on fixed threshold or simple rule to judge the fault position, which is difficult to adapt to the complex nonlinear change of fault after the access of distributed photovoltaic. The existing technology lacks quantitative evaluation of the fault occurrence probability of different line units under specific electrical parameter combination, resulting in that the positioning process relies on experience and is easily disturbed.

[0052] The method provided in the embodiments of the present application comprises the following steps S30:

[0053] The line units between the upstream key switch and the downstream adjacent key switch in the sequence of line units are divided as a plurality of implied abnormal line units.

[0054] The upstream abnormal switch electrical parameter and the compensated downstream abnormal switch electrical parameter are input into a fault unit predictor to predict and output a plurality of line unit fault rates of the plurality of implied abnormal line units.

[0055] The training step of the fault unit predictor comprises the following steps:

[0056] According to the power grid fault monitoring data in the historical time, the upstream abnormal switch electrical parameter set and the downstream abnormal switch electrical parameter set of the sample fault occurring multiple times are collected, and the proportion of the fault occurring in different positions of the line unit in the multiple times of fault occurring is collected, and a plurality of sample line unit probability sets are labeled.

[0057] Based on machine learning, a fault unit predictor is constructed.

[0058] The fault unit predictor is supervised trained by using the upstream abnormal switch electrical parameter set, the downstream abnormal switch electrical parameter set and the plurality of sample line unit probability sets, and is completed after training test convergence.

[0059] In the embodiments of the present application, the line units between the upstream key switch and the downstream adjacent key switch in the line unit sequence are divided as a plurality of implicit abnormal line units. Specifically, the line units between the upstream key switch and the downstream adjacent key switch in the line unit sequence are searched and divided as a plurality of implicit abnormal line units.

[0060] According to the power grid fault monitoring data in the historical time, the upstream abnormal switch electrical parameter set and the downstream abnormal switch electrical parameter set of the samples in which faults occur multiple times are collected, and the proportion of faults occurring in different position line units in the multiple times of faults is collected, the line units are labeled, the labeling content is the proportion of faults, and the plurality of sample line unit probability sets are obtained.

[0061] Machine learning is used to construct a fault unit predictor, for example, a three-layer structure is used, the input layer is used to receive the upstream abnormal switch electrical parameters and the downstream abnormal switch electrical parameters, the hidden layer uses 16 nodes and uses the ReLU function for activation, the output layer outputs the predicted line unit probability, and the loss function uses the cross-entropy function.

[0062] The fault unit predictor is supervised trained by using the upstream abnormal switch electrical parameter set, the downstream abnormal switch electrical parameter set and the plurality of sample line unit probability sets, and is completed after training test convergence, for example, the line unit probability accuracy error of the output is within ±5% when the upstream abnormal switch electrical parameters and the downstream abnormal switch electrical parameters are input, that is, the fault unit predictor training is completed.

[0063] The upstream abnormal switch electrical parameters and the compensated downstream abnormal switch electrical parameters are input into the fault unit predictor, and the plurality of line unit fault rates of the plurality of implicit abnormal line units are obtained by prediction output. For example, the line unit numbered 1002 has a predicted fault rate of 60%, and the line unit numbered 2031 has a predicted fault rate of 20%.

[0064] The pre-trained fault unit predictor is used in the present application, the upstream abnormal switch electrical parameters and the compensated downstream abnormal switch electrical parameters are input, and the line unit fault rate is output. The predictor learns the deep correlation between the electrical parameter mode and the fault position in the historical fault data through machine learning, can automatically capture the complex mode of fault features, realizes the prediction from the electrical parameters to the fault probability, and significantly improves the objectivity and accuracy of the candidate fault section evaluation.

[0065] S40: According to the multiple implicit abnormal line units, fault electrical parameter prediction is performed to obtain multiple predicted upstream abnormal switch electrical parameters and multiple predicted downstream abnormal switch electrical parameters, similarity of each group of predicted upstream abnormal switch electrical parameters and predicted downstream abnormal switch electrical parameters with the upstream abnormal switch electrical parameters and the compensated downstream abnormal switch electrical parameters is calculated, and a fault line unit is selected as a fault positioning result by combining the multiple line unit failure rates.

[0066] Only relying on fault probability prediction still has uncertainty, and a single probability value may not be sufficient to distinguish similar candidate units. The existing method often ignores the matching degree verification of the theoretical fault characteristics and the actual monitoring value, or does not fuse the probability information when verifying, which is easy to cause the positioning result to deviate from the actual situation.

[0067] The step S40 in the method provided in the embodiment of the application comprises:

[0068] The line unit number of each implicit abnormal line unit is input into a fault electrical parameter predictor to obtain multiple predicted upstream abnormal switch electrical parameters and multiple predicted downstream abnormal switch electrical parameters.

[0069] The construction step of the fault electrical parameter predictor comprises:

[0070] According to the fault monitoring data of the power distribution network in the historical time, a sample line unit number set, a sample upstream abnormal switch electrical parameter set and a sample downstream abnormal switch electrical parameter set when different line units fail are collected.

[0071] The sample line unit number set, the sample upstream abnormal switch electrical parameter set and the sample downstream abnormal switch electrical parameter set are used to construct the fault electrical parameter predictor.

[0072] The average similarity of each group of predicted upstream abnormal switch electrical parameters and predicted downstream abnormal switch electrical parameters with the upstream abnormal switch electrical parameters and the compensated downstream abnormal switch electrical parameters is calculated to obtain multiple fault similarities.

[0073] According to the multiple fault similarities and the multiple line unit failure rates, a fault line unit is selected as a fault positioning result by optimization.

[0074] According to the multiple fault similarities and the multiple line unit failure rates, a fault line unit is selected as a fault positioning result by optimization, which comprises:

[0075] According to the multiple fault similarities and the multiple line unit failure rates, multiple comprehensive fault confidences are calculated.

[0076] The fault line unit with the maximum comprehensive fault confidence is selected as the fault positioning result.

[0077] In the embodiments of the present application, according to the power distribution network fault monitoring data in the historical time, the sample line unit number set, the sample upstream abnormal switch electrical parameter set and the sample downstream abnormal switch electrical parameter set when different line units appear faults are collected.

[0078] A neural network is adopted. A fault electrical parameter predictor is constructed. Exemplarily, a three-layer neural network fault electrical parameter predictor is constructed, wherein 1 node is adopted in the input layer, the line unit number is input, 16 nodes are adopted in the hidden layer, the ReLU function is activated, 2 nodes are adopted in the output layer, the predicted upstream abnormal switch electrical parameter and the predicted downstream abnormal switch electrical parameter are output, and the mean square error is adopted as the loss function. The sample line unit number set, the sample upstream abnormal switch electrical parameter set and the sample downstream abnormal switch electrical parameter set when different line units appear faults are adopted to supervise the training of the constructed fault electrical parameter predictor until convergence, for example, the accuracy rate of the input line unit number, the output predicted upstream abnormal switch electrical parameter and the output predicted downstream abnormal switch electrical parameter is more than 90%, that is, the training of the fault electrical parameter predictor is completed.

[0079] The line unit number of each hidden abnormal line unit is input into the fault electrical parameter predictor, and a plurality of predicted upstream abnormal switch electrical parameters and a plurality of predicted downstream abnormal switch electrical parameters are obtained by prediction output. The predicted electrical parameters reflect the possible fault electrical parameter situation when the line unit appears faults.

[0080] The average similarity of each group of predicted upstream abnormal switch electrical parameters and predicted downstream abnormal switch electrical parameters and upstream abnormal switch electrical parameters and compensated downstream abnormal switch electrical parameters is calculated to obtain a plurality of fault similarities. For example, the upstream abnormal similarity = 1- | predicted upstream abnormal switch electrical parameter - upstream abnormal switch electrical parameter | ÷ [(predicted upstream abnormal switch electrical parameter + upstream abnormal switch electrical parameter) ÷ 2], the downstream abnormal similarity = 1- | predicted downstream abnormal switch electrical parameter - compensated downstream abnormal switch electrical parameter | ÷ [(predicted downstream abnormal switch electrical parameter + compensated downstream abnormal switch electrical parameter) ÷ 2], for example, the predicted upstream abnormal switch electrical parameter is 100A, the upstream abnormal switch electrical parameter is 80A, then the upstream abnormal similarity = 1- | 100-80 | ÷ [(100+80) ÷ 2] = 0.78, the predicted downstream abnormal switch electrical parameter is 90A, the compensated downstream abnormal switch electrical parameter is 80A, then the downstream abnormal similarity = 1- | 90-80 | ÷ [(90+80) ÷ 2] = 0.88. The average similarity = (upstream abnormal similarity + downstream abnormal similarity) ÷ 2, for example, the upstream abnormal similarity is 0.78, the downstream abnormal similarity is 0.88, then the average similarity = (0.78+0.88) ÷ 2 = 0.83. The higher the similarity is, the closer the upstream and downstream electrical parameters of the line unit are to the possible upstream and downstream electrical parameters when the fault occurs, and the greater the probability of the line unit fault is.

[0081] According to the plurality of fault similarities and the plurality of line unit fault rates, a plurality of comprehensive fault confidence degrees are calculated. The comprehensive fault confidence degree = the fault similarity of the line unit × the fault rate of the line unit, for example, the fault similarity of the line unit is 0.83, and the fault rate is 0.80, then the comprehensive fault confidence degree = 0.83 × 0.80 = 0.664.

[0082] The fault line unit with the maximum comprehensive fault confidence degree is selected as the fault positioning result.

[0083] The predicted upstream and downstream abnormal switch electrical parameters of each line unit are generated by the fault electrical parameter predictor, the similarity of the predicted upstream and downstream abnormal switch electrical parameters and the upstream abnormal switch electrical parameters and the compensated downstream abnormal switch electrical parameters is calculated, the comprehensive fault confidence degree is calculated, and the optimal fault line unit is selected. The double verification mechanism combines the statistical advantage of probability prediction and the physical consistency of feature matching, effectively overcomes the limitations of a single method, significantly improves the robustness and reliability of the fault positioning result, and ensures accurate locking of the real fault point.

[0084] Embodiment two, as shown in Figure 2 based on the same inventive concept of the fault positioning processing method of the distributed photovoltaic power distribution network provided in embodiment one, the embodiment of the present application further provides a fault positioning processing system of a distributed photovoltaic power distribution network, comprising:

[0085] The digital modeling module 100 is configured to perform digital modeling on a target power distribution network containing distributed photovoltaics, and to perform line unit division to obtain a line unit sequence, wherein the target power distribution network includes a plurality of key switches arranged at a plurality of positions.

[0086] The compensation parameter calculation module 200 is configured to, when detecting that any one of the switch electrical parameters at the key switches is greater than or equal to an electrical parameter threshold, obtain an upstream abnormal switch electrical parameter and a downstream abnormal switch electrical parameter of a downstream adjacent key switch, and calculate a compensated downstream abnormal switch electrical parameter.

[0087] The fault rate prediction module 300 is configured to divide a plurality of implicit abnormal line units in the line unit sequence, and to perform line unit fault rate prediction according to the upstream abnormal switch electrical parameter and the compensated downstream abnormal switch electrical parameter to obtain a plurality of line unit fault rates.

[0088] The fault location module 400 is configured to perform fault electrical parameter prediction according to the plurality of implicit abnormal line units to obtain a plurality of predicted upstream abnormal switch electrical parameters and a plurality of predicted downstream abnormal switch electrical parameters, to calculate similarity with the upstream abnormal switch electrical parameter and the compensated downstream abnormal switch electrical parameter respectively, and to combine the plurality of line unit fault rates to optimally select a fault line unit as a fault location result.

[0089] In one embodiment, the digital modeling module 100 is further configured to:

[0090] Obtain power grid feature information of a target power distribution network, perform digital modeling, and obtain a power distribution network model, wherein the target power distribution network contains distributed photovoltaic power supply and includes a plurality of key switches arranged at a plurality of positions.

[0091] Perform line unit division on lines in the power distribution network model to obtain a line unit sequence.

[0092] In one embodiment, the compensation parameter calculation module 200 is further configured to:

[0093] Monitor switch electrical parameters at the plurality of key switches, and when any one of the switch electrical parameters is greater than or equal to an electrical parameter threshold, obtain an upstream key switch and take the switch electrical parameter as an upstream abnormal switch electrical parameter.

[0094] Obtain a switch electrical parameter of a downstream adjacent key switch in the current direction as a downstream abnormal switch electrical parameter.

[0095] According to distributed photovoltaic operation and maintenance data, index a distributed photovoltaic electrical parameter corresponding to the downstream adjacent key switch.

[0096] Subtracting the distributed photovoltaic electrical parameter from the downstream abnormal switching electrical parameter obtains a compensated downstream abnormal switching electrical parameter.

[0097] In one embodiment, the fault rate prediction module 300 is further configured to:

[0098] Divide the line units between the upstream key switch and the downstream adjacent key switch in the line unit sequence as a plurality of implied abnormal line units.

[0099] Input the upstream abnormal switching electrical parameter and the compensated downstream abnormal switching electrical parameter into the fault unit predictor to obtain a plurality of line unit fault rates of the plurality of implied abnormal line units.

[0100] The training step of the fault unit predictor comprises:

[0101] According to the power grid fault monitoring data in the historical time, a set of sample upstream abnormal switching electrical parameters and a set of sample downstream abnormal switching electrical parameters of multiple fault occurrences are collected, and the proportion of fault occurrences of line units at different positions in multiple fault occurrences is collected to obtain a plurality of sample line unit probability sets.

[0102] Based on machine learning, a fault unit predictor is constructed.

[0103] The sample upstream abnormal switching electrical parameter set, the sample downstream abnormal switching electrical parameter set, and the plurality of sample line unit probability sets are used to supervise the training of the fault unit predictor, which is completed after training test convergence.

[0104] In one embodiment, the fault location module 400 is further configured to:

[0105] Input the line unit number of each implied abnormal line unit into the fault electrical parameter predictor to obtain a plurality of predicted upstream abnormal switching electrical parameters and a plurality of predicted downstream abnormal switching electrical parameters.

[0106] The construction step of the fault electrical parameter predictor comprises:

[0107] According to the distribution network fault monitoring data in the historical time, a set of sample line unit numbers, a set of sample upstream abnormal switching electrical parameters, and a set of sample downstream abnormal switching electrical parameters when different line units fail are collected.

[0108] The set of sample line unit numbers, the set of sample upstream abnormal switching electrical parameters, and the set of sample downstream abnormal switching electrical parameters are used to construct a fault electrical parameter predictor.

[0109] The similarity between each set of predicted upstream abnormal switch electrical parameters and predicted downstream abnormal switch electrical parameters and the average of the upstream abnormal switch electrical parameters and the compensated downstream abnormal switch electrical parameters is calculated to obtain a plurality of fault similarities.

[0110] According to the plurality of fault similarities and a plurality of line unit failure rates, a fault line unit is obtained by optimization selection as a fault location result.

[0111] According to the plurality of fault similarities and a plurality of line unit failure rates, a fault line unit is obtained by optimization selection as a fault location result, including:

[0112] According to the plurality of fault similarities and a plurality of line unit failure rates, a plurality of comprehensive fault confidence degrees are calculated.

[0113] The fault line unit with the maximum comprehensive fault confidence degree is selected by screening selection as a fault location result.

[0114] Embodiment three, as Figure 3 shown, the embodiment of the present application also provides a non-transitory computer readable storage medium 500, and the non-transitory computer readable storage medium stores a computer program 511, and the computer program 511 is executed by a processor to realize the fault location method of the distributed photovoltaic power distribution network as in embodiment one.

[0115] To sum up, the embodiments of the present application have at least the following technical effects:

[0116] The present application provides a fault location method, system and storage medium of a distributed photovoltaic power distribution network, by establishing an accurate digital model and dividing a line unit sequence, when detecting abnormal key switch electrical parameters, introducing a compensation mechanism for the downstream adjacent key switch abnormal parameters of the distributed photovoltaic, effectively removing the pollution of the photovoltaic power supply to the fault characteristic signal, and restoring the real fault electrical quantity characteristics; then, based on the compensated upstream and downstream abnormal parameters, using a pre-trained fault unit predictor to predict the potential fault line unit, and combining the similarity verification of the theoretical parameters generated by the fault electrical parameter predictor and the actual monitoring parameters, the fault probability and the feature matching degree are fused for optimization decision, which significantly improves the accuracy and robustness of fault location. Compared with the traditional method, the technical scheme provided by the present application significantly overcomes the change of fault current characteristics caused by the grid-connected distributed photovoltaic, and avoids the misjudgment or omission of fault location caused by continuous photovoltaic power supply.

[0117] The present application achieves the technical effect of accurately and reliably identifying the fault section of the power distribution network containing distributed photovoltaic, guiding the operation and maintenance personnel to improve the fault handling efficiency and the safety of power grid operation.

[0118] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0119] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0120] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A method for fault location and processing in a distributed photovoltaic distribution network, characterized in that: The method comprises: Digitally modeling a target distribution network including distributed photovoltaics and dividing it into line units to obtain a line unit sequence, wherein the target distribution network includes a plurality of key switches configured at a plurality of locations; When it is detected that the switch electrical parameters at any key switch are greater than or equal to the electrical parameter threshold, the electrical parameters of the upstream abnormal switch and the electrical parameters of the downstream abnormal switch adjacent to the downstream key switch are obtained, and the electrical parameters of the downstream abnormal switch are calculated to compensate, including: Monitor switch electrical parameters at multiple key switches, and when any switch electrical parameter is greater than or equal to an electrical parameter threshold, obtain the upstream key switch and use the switch electrical parameter as the upstream abnormal switch electrical parameter; Obtaining electrical parameters of a key switch adjacent to the downstream side of the current flow as electrical parameters of the downstream abnormal switch; Based on the distributed photovoltaic operation and maintenance data, index and obtain the distributed photovoltaic electrical parameters corresponding to the downstream adjacent key switches; Subtracting the distributed photovoltaic electrical parameters from the electrical parameters of the downstream abnormal switch to obtain compensated electrical parameters of the downstream abnormal switch; Dividing the line unit sequence to obtain a plurality of implicitly abnormal line units, and performing line unit failure rate prediction based on the electrical parameters of the upstream abnormal switch and the electrical parameters of the compensated downstream abnormal switch to obtain a plurality of line unit failure rates; Fault electrical parameters are predicted based on multiple implicit abnormal line units to obtain multiple predicted electrical parameters of upstream abnormal switches and multiple predicted electrical parameters of downstream abnormal switches. Similarities with the electrical parameters of the upstream abnormal switches and the electrical parameters of the compensated downstream abnormal switches are calculated to obtain multiple fault similarities. Combined with the multiple line unit failure rates, the faulty line unit is optimally selected as the fault location result, including: According to multiple fault similarities and multiple line unit failure rates, multiple comprehensive fault confidences are calculated and obtained; The fault line unit with the largest comprehensive fault confidence is selected as the fault location result.

2. The fault location processing method for a distributed photovoltaic distribution network according to claim 1, characterized in that: Digitally model the target distribution network including distributed photovoltaics and divide it into line units to obtain a line unit sequence, including: Obtaining grid characteristic information of a target distribution network, performing digital modeling, and obtaining a distribution network model, wherein the target distribution network includes distributed photovoltaic power supply and includes multiple key switches configured at multiple locations; The lines in the distribution network model are divided into line units to obtain a line unit sequence.

3. The fault location processing method for a distributed photovoltaic power distribution network according to claim 1, characterized in that: A plurality of implicit abnormal line units are obtained by dividing the line unit sequence, and a line unit failure rate prediction is performed based on the electrical parameters of the upstream abnormal switch and the electrical parameters of the compensated downstream abnormal switch to obtain a plurality of line unit failure rates, including: dividing the line units between the upstream key switch and the downstream adjacent key switch in the line unit sequence as a plurality of implicit abnormal line units; The electrical parameters of the upstream abnormal switch and the electrical parameters of the compensated downstream abnormal switch are input into a fault unit predictor, and a prediction output is obtained to obtain multiple line unit failure rates of multiple implicit abnormal line units.

4. The fault location processing method for a distributed photovoltaic power distribution network according to claim 3, characterized in that: The training step of the fault unit predictor includes: Based on the historical power grid fault monitoring data, the electrical parameter sets of upstream abnormal switches and downstream abnormal switches with multiple faults are collected. The proportion of line units with faults at different locations in the multiple faults is collected, and the probability sets of multiple line units are obtained by annotation. Build a fault unit predictor based on machine learning; The fault unit predictor is supervised trained using the sample upstream abnormal switch electrical parameter set, the sample downstream abnormal switch electrical parameter set and a plurality of sample line unit probability sets, and the training is completed after the training and testing converge.

5. The fault location processing method for a distributed photovoltaic power distribution network according to claim 1, characterized in that: Fault electrical parameters are predicted based on multiple implicit abnormal line units to obtain multiple predicted electrical parameters of upstream abnormal switches and multiple predicted electrical parameters of downstream abnormal switches. Similarities with the electrical parameters of the upstream abnormal switches and the electrical parameters of the compensated downstream abnormal switches are calculated respectively. Combined with the failure rates of the multiple line units, the faulty line unit is optimally selected and obtained as the fault location result, including: Inputting the line unit number of each implicit abnormal line unit into a fault electrical parameter predictor, and predicting output to obtain a plurality of predicted upstream abnormal switch electrical parameters and a plurality of predicted downstream abnormal switch electrical parameters; Calculating the average similarity of each set of predicted upstream abnormal switch electrical parameters and predicted downstream abnormal switch electrical parameters with the upstream abnormal switch electrical parameters and compensated downstream abnormal switch electrical parameters to obtain multiple fault similarities; According to multiple fault similarities and multiple line unit failure rates, the faulty line unit is optimally selected and obtained as the fault location result.

6. The fault location processing method for a distributed photovoltaic power distribution network according to claim 5, characterized in that: The steps of constructing the fault electrical parameter predictor include: Based on the distribution network fault monitoring data in the historical period, a sample line unit number set, a sample upstream abnormal switch electrical parameter set, and a sample downstream abnormal switch electrical parameter set are collected when faults occur in different line units; A fault electrical parameter predictor is constructed by using the sample line unit number set, the sample upstream abnormal switch electrical parameter set, and the sample downstream abnormal switch electrical parameter set.

7. A fault location and processing system for a distributed photovoltaic distribution network, characterized in that: A system for implementing a fault location processing method for a distributed photovoltaic distribution network according to any one of claims 1 to 6, comprising: a digital modeling module, configured to digitally model a target distribution network including distributed photovoltaics and divide the network into line units to obtain a line unit sequence, wherein the target distribution network includes a plurality of key switches configured at a plurality of locations; The compensation parameter calculation module is used to obtain the electrical parameters of the upstream abnormal switch and the electrical parameters of the downstream abnormal switch of the downstream adjacent key switch when the switch electrical parameters at any key switch are detected to be greater than or equal to the electrical parameter threshold, and calculate the electrical parameters of the compensated downstream abnormal switch, including: Monitor switch electrical parameters at multiple key switches, and when any switch electrical parameter is greater than or equal to an electrical parameter threshold, obtain the upstream key switch and use the switch electrical parameter as the upstream abnormal switch electrical parameter; Obtaining electrical parameters of a key switch adjacent to the downstream side of the current flow as electrical parameters of the downstream abnormal switch; Based on the distributed photovoltaic operation and maintenance data, index and obtain the distributed photovoltaic electrical parameters corresponding to the downstream adjacent key switches; Subtracting the distributed photovoltaic electrical parameters from the electrical parameters of the downstream abnormal switch to obtain compensated electrical parameters of the downstream abnormal switch; a failure rate prediction module, configured to obtain a plurality of implicitly abnormal line units by dividing the line unit sequence, and predicting the line unit failure rate based on the electrical parameters of the upstream abnormal switch and the electrical parameters of the compensated downstream abnormal switch, to obtain a plurality of line unit failure rates; A fault location module is configured to predict fault electrical parameters based on multiple implicitly abnormal line units, obtain multiple predicted electrical parameters of upstream abnormal switches and multiple predicted electrical parameters of downstream abnormal switches, calculate similarities with the electrical parameters of the upstream abnormal switches and the electrical parameters of the compensated downstream abnormal switches, obtain multiple fault similarities, and optimize and select the faulty line unit based on the failure rates of the multiple line units as the fault location result, including: According to multiple fault similarities and multiple line unit failure rates, multiple comprehensive fault confidences are calculated and obtained; The fault line unit with the largest comprehensive fault confidence is selected as the fault location result.

8. A non-transitory computer-readable storage medium, characterized in that The storage medium stores a computer program, and when the computer program is executed by the processor, the method for fault location processing of a distributed photovoltaic distribution network according to any one of claims 1 to 6 is implemented.

Images