A fault analysis and location system for distribution network

Abstract

The invention discloses a fault analysis and location system for a distribution network, relates to the technical field of intelligent power distribution, and solves the technical problems that the prior art has low efficiency in locating faults in complex distribution networks and is difficult to accurately identify the fault types of distribution networks; the invention collects the current flowing through the intelligent switch on the distribution link; when the current has an abnormal change, the current balance of the adjacent topological nodes downstream of the intelligent switch is analyzed in real time to determine whether there is a line fault; that is, the recorded wave data of the faulty line at the same moment is uploaded; the invention can avoid the time delay caused by the data transmission process, and locate the fault according to the current balance of the topological nodes at the same moment; the invention extracts the recorded wave data of the faulty line, extracts features of the electrical data before and after the line fault in the recorded wave data, integrates and generates a fault feature sequence, and determines the fault type in combination with a fault judgment model; the invention extracts the features of the recorded wave data before and after based on the same moment, and determines the fault type efficiently and accurately in combination with the fault judgment model.

Description

Technical Field

[0001] The invention belongs to the field of intelligent power distribution, relates to a distribution network fault location technology, and specifically is a distribution network fault analysis and location system. Background Art

[0002] The distribution network has the characteristics of multiple voltage levels, complex network structure, various equipment types, multiple and wide operating points, and relatively poor safety environment. Real-time monitoring and mobile operation and maintenance of the distribution network can quickly locate the fault location once a problem occurs in the distribution network, which can increase the service life of the distribution network and ensure safe and stable power supply of the distribution network. Therefore, it is very necessary to analyze the real-time monitoring data after a line fault occurs.

[0003] In the fault location of the distribution network, the electrical data of each distribution line in the distribution network is generally collected in real time, and the distribution line is judged whether a fault occurs by comparing each electrical data with the corresponding threshold value or the difference between the waveform change of each electrical data and the corresponding standard waveform. When locating the fault of the complex distribution line in the distribution network, the existing technology needs to process a large amount of data, which affects the positioning efficiency. In addition, due to the complexity of the distribution line, it is difficult to accurately identify the fault type, and it is impossible to perform timely emergency repair and maintenance on the distribution network fault; therefore, a fault analysis and location system for the distribution network is urgently needed. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a fault analysis and location system for distribution network, which is used to solve the technical problems that the prior art has low efficiency in locating faults in complex distribution networks, makes it difficult to accurately identify the fault types of distribution networks, and results in the inability to perform timely emergency repairs and maintenance.

[0005] To achieve the above-mentioned object, the first aspect of the present invention provides a fault analysis and location system for a distribution network, comprising a central control module and an intelligent acquisition module connected thereto;

[0006] The central control module establishes a power grid topology model; marks the distribution lines between adjacent substations in the power grid topology model as distribution links, and marks each power branch in the distribution link as a topology link; wherein the distribution branch or the distribution branch is provided with an intelligent switch, and the substation unit includes a transformer or a substation;

[0007] The intelligent acquisition module collects the current flowing through the intelligent switch on the distribution link; when the current changes abnormally, it analyzes the current balance of the adjacent topological nodes downstream of the intelligent switch in real time to determine whether there is a line fault; if yes, it uploads the recorded waveform data of the faulty line at the same time; the topological node is the intersection of the distribution link and the topological link;

[0008] The central control module extracts the recorded data, extracts features from the electrical data before and after the line fault in the recorded data, integrates and generates a fault feature sequence, and determines the fault type based on the fault judgment model; among them, the fault judgment model is established based on the artificial intelligence model, and the electrical data includes voltage and current.

[0009] Preferably, the central control module communicates and / or is electrically connected with the intelligent acquisition module; and the intelligent acquisition module is associated with the intelligent switch and configured in the topological link of the distribution network;

[0010] The intelligent acquisition module collects electrical data through an associated intelligent switch; and the intelligent switch is disconnected when a line fault occurs in the corresponding topological link and the fault current is greater than a current threshold.

[0011] Preferably, the central control module establishes a power grid topology model, including:

[0012] Obtaining distribution network laying data through smart terminals; wherein the smart terminals include mobile phones or computers;

[0013] The distribution network laying data is used to identify the substation units and the corresponding distribution links, as well as the topological nodes and topological links in the distribution links. After integration, they are visualized and a power grid topology model is generated.

[0014] Preferably, the intelligent acquisition module acquires the current flowing through the intelligent switch on the power distribution link and determines whether there is an abnormal change in the current, including:

[0015] Collect the current flowing through the smart switch in real time;

[0016] Compare the characteristic data of the current with the corresponding standard characteristics to see if they are consistent; if yes, determine that no abnormality has occurred; if no, determine that current abnormality has occurred; wherein the characteristic data includes waveform or effective value.

[0017] Preferably, the real-time analysis of the current balance of the adjacent topological nodes downstream of the intelligent switch to determine whether there is a line fault includes:

[0018] When there is an abnormal change in the current, several topological nodes corresponding to the downstream of the intelligent switch are identified; wherein the upstream and downstream of the intelligent switch are determined by the current flow direction;

[0019] Based on Kirchhoff's first law, it is verified whether the current of the downstream topological node is balanced at the same time, and then the fault line is determined according to the current balance verification result.

[0020] Preferably, the verifying whether the current of the downstream topological node is balanced at the same time based on Kirchhoff's first law includes:

[0021] Identify the number of downstream topological nodes based on the power grid topology model, and then determine the verification interval;

[0022] The downstream topological nodes are continuously divided into several node groups according to the verification interval; the most downstream topological node in each node group is used as the target node, and the remaining topological nodes are converted into topological links of the target node;

[0023] Kirchhoff's first law is used to verify whether the current of the target node is balanced at the same time; if yes, the next target node is verified; if not, the topological nodes in the node group corresponding to the target node are verified one by one.

[0024] Preferably, the feature extraction of electrical data before and after the line fault in the recorded data and integration to generate a fault feature sequence includes:

[0025] Extracting electrical characteristics of electrical data before and after the line fault, and the difference between the electrical characteristics of the electrical data before and after the fault from the recorded data; wherein the electrical characteristics include waveform, phase or effective value;

[0026] The electrical characteristics before and after the line fault and the differences in electrical characteristics are integrated to generate a fault feature sequence.

[0027] Preferably, the fault judgment model is established based on an artificial intelligence model, including:

[0028] Acquire standard training data; wherein the standard training data includes standard input data consistent with the content attributes of the fault feature sequence, and standard output data indicating the fault type;

[0029] The artificial intelligence model constructed by training with standard training data is marked as a fault judgment model after training is completed; wherein the artificial intelligence model includes a BP neural network model or a RBF neural network model.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. The present invention collects the current flowing through the intelligent switch on the distribution link; when there is an abnormal change in the current, the current balance of the adjacent topological nodes downstream of the intelligent switch is analyzed in real time to determine whether there is a line fault; if yes, the recorded wave data of the faulty line is uploaded at the same time; the present invention can avoid the time delay caused by the data transmission process and locate the fault according to the current balance of the topological nodes at the same time.

[0032] 2. The present invention extracts the recorded data of the faulty line, performs feature extraction on the electrical data before and after the line fault in the recorded data, integrates and generates a fault feature sequence, and determines the fault type in combination with a fault judgment model; the present invention extracts the features of the recorded data before and after based on the same moment, and determines the fault type efficiently and accurately in combination with a fault judgment model. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 paying creative work.

[0034] Figure 1 It is a schematic diagram of the system principle of the present invention;

[0035] Figure 2 It is a schematic diagram of the method steps of the present invention;

[0036] Figure 3 It is a schematic diagram of the connection between the power distribution link and the topology link of the present invention;

[0037] Figure 4 It is a schematic diagram of node group segmentation of the present invention. DETAILED DESCRIPTION

[0038] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] See also Figure 1-Figure 2 The first aspect of the present invention provides a fault analysis and positioning system for a distribution network, including a central control module and an intelligent acquisition module connected thereto; the central control module establishes a power grid topology model; the distribution lines between adjacent substations in the power grid topology model are marked as distribution links, and each power branch in the distribution link is marked as a topology link; the intelligent acquisition module collects the current flowing through the intelligent switch on the distribution link; when the current changes, the current balance of the adjacent topological nodes downstream of the intelligent switch is analyzed in real time to determine whether there is a line fault; if yes, the recorded data of the faulty line at the same time is uploaded; the central control module extracts the recorded data, performs feature extraction on the electrical data before and after the line fault in the recorded data, integrates and generates a fault feature sequence, and determines the fault type in combination with the fault judgment model.

[0040] In the present invention, the central control module communicates and / or is electrically connected with the intelligent acquisition module; and the intelligent acquisition module is associated with the intelligent switch and configured in the topological link of the distribution network; the intelligent acquisition module collects electrical data through the associated intelligent switch; and the intelligent switch is disconnected when a line fault occurs in the topological link to which it belongs and the fault current is greater than the current threshold.

[0041] The central control module is mainly responsible for data processing and analysis, including building a fault judgment model and judging the fault type based on the electrical data of the faulty line. It mainly obtains relevant data from the intelligent acquisition module and the intelligent terminal. The intelligent terminal transmits some basic data of the distribution network to the central control module so that the grid topology model can be built. The intelligent acquisition module is associated with the intelligent switch and collects electrical data flowing through the intelligent switch.

[0042] It is worth noting that in order to ensure timely and accurate fault location, the intelligent acquisition modules communicate with each other to ensure fast data transmission, that is, the intelligent acquisition modules can quickly extract data from other intelligent acquisition modules. At the same time, the intelligent acquisition modules also need to have certain data storage functions to ensure the possibility of "simultaneous" data analysis.

[0043] In the present invention, the distribution lines between adjacent substations are marked as distribution links, and each power branch in the distribution link is marked as a topological link. The distribution link and the topological link are marked to normalize the distribution network from whole to zero, so that Kirchhoff's first law can be better used to quickly locate the fault line in fault location. Moreover, the intersection of the distribution link and the topological link is used as a topological node. Intelligent switches are set in the distribution link and the topological link, and intelligent switches are set between adjacent topological nodes on the distribution link.

[0044] See also Figure 3 The substation in the present invention includes a transformer or a substation. For example, the distribution line between transformer A and the transformer B directly connected to it is a distribution link, the power line drawn from the distribution link is a topological link, and the lead-out point is a topological node. The main purpose of determining the distribution line by the substation is to meet the application conditions of Kirchhoff's first law. Therefore, the substation can be extended. For example, when electricity is introduced from a substation to a resident's home, the grounding end of the distribution line can also be used as a substation, that is, a distribution link is formed between the transformer and the grounding end, and the residential power line is used as a topological link.

[0045] In a preferred embodiment, the central control module establishes a power grid topology model, including: obtaining distribution network laying data through an intelligent terminal; identifying substation units and corresponding distribution links through the distribution network laying data, and identifying topological nodes and topological links in the distribution links, integrating them for visual display, and generating a power grid topology model.

[0046] Identify the distribution lines, substations, power equipment, etc. in the distribution network laying data, visualize the distribution lines through GIS technology, and generate a power grid topology model. In the power grid topology model, the distribution links and topology links are marked according to their definitions, and the topology nodes are marked in the power grid topology model. The subsequent fault location process can be updated in real time in the power grid topology model. For example, when multiple topological nodes need to be verified for current balance, the verification process is displayed in real time in the power grid topology model, which can help maintenance personnel understand the progress and help maintenance personnel reduce the scope of the fault in a short time. It is understandable that the power grid topology model can be updated and displayed in a timely manner on the maintenance personnel's smart terminal.

[0047] When locating a distribution network fault, first roughly determine whether a fault has occurred in the distribution network. The intelligent acquisition module collects the current flowing through the intelligent switch on the distribution link and determines whether the current has changed, including: real-time collection of the current flowing through the intelligent switch; comparison of the current characteristic data with the corresponding standard characteristics; if yes, determine that no abnormality has occurred; if no, determine that the current has changed abnormally.

[0048] Each distribution link in the power grid topology model is an independent and smallest fault monitoring unit. By analyzing the abnormal changes of the current flowing through the intelligent switch on the distribution link, it is possible to quickly determine whether the distribution link is faulty; if it is faulty, current balance analysis is performed on the topological nodes on the distribution link to locate the fault.

[0049] In this embodiment, judging whether the current is abnormal is actually judging the characteristic data of the current, such as the waveform or the effective value. When the waveform of the current is consistent with the normal waveform, and the effective value deviates less than the effective value under normal conditions, it is judged that the current has not changed abnormally; otherwise, it is judged that the current is abnormal, that is, there must be current imbalance in the topological nodes in the corresponding distribution link at the same time.

[0050] The simultaneous moment or the same time in the present invention mainly refers to the correspondence between the line fault moment and the data collection moment. If the current is analyzed to be abnormal, the current balance analysis needs to be performed next, and the current data used in the current balance analysis is consistent with the collection time of the abnormal current data, so as to ensure the accuracy of the analysis and judgment, and avoid the delay caused by data transmission affecting the data analysis results.

[0051] Then it is necessary to further determine whether the current is faulty. Real-time analysis of the current balance of the adjacent topological nodes downstream of the smart switch is performed to determine whether there is a line fault, including: when there is an abnormal change in the current, several topological nodes downstream of the corresponding smart switch are identified; based on Kirchhoff's first law, the current of the downstream topological nodes is verified to be balanced at the same time, and then the faulty line is determined based on the current balance verification results.

[0052] Kirchhoff's first law can be applied in the present invention and can be summarized as the current flowing into the topological node is equal to the current flowing out of the topological node. Based on the current change analysis, it can be determined whether the distribution line is faulty. If it is, the downstream topological nodes need to be analyzed one by one, that is, the current balance analysis is performed one by one. Once the current imbalance occurs, it can be determined that the topological link corresponding to the topological node is faulty, or part of the distribution link is faulty. Please refer to Figure 3 , through the intelligent switch directly connected to substation A, it is judged that the current in the distribution link between substation A and substation B changes abnormally. Figure 3 It only needs to check whether the topological nodes (white filled points) in the figure satisfy the current balance, that is, the current balance is determined by the currents of the three smart switches (black filled points) around the topological nodes.

[0053] In the present invention, the upstream and downstream of the intelligent switch are determined by the current flow direction. If the current flow direction is positive, the current flows through topological node 1, flows through the intelligent switch, and then flows through topological node 2, then topological node 1 is upstream relative to the intelligent switch, and topological node 2 is downstream relative to the intelligent switch.

[0054] By analyzing whether the current of the topological nodes is balanced one by one, the fault line can be located, and the fault location of the distribution network is completed at this moment. However, if there are too many topological nodes corresponding to the same distribution line, analyzing them one by one will seriously affect the efficiency of fault location.

[0055] See also Figure 4 In another optional embodiment, the current of the downstream topological nodes is verified to be balanced at the same time based on Kirchhoff's first law, including: identifying the number of downstream topological nodes based on the power grid topology model, and then determining the verification interval; continuously dividing the downstream topological nodes into a number of node groups according to the verification interval; taking the most downstream topological node in each node group as the target node, and converting the remaining topological nodes into topological links of the target node; verifying whether the current of the target node is balanced at the same time through Kirchhoff's first law; if yes, verifying the next target node; if not, verifying the topological nodes in the node group corresponding to the target node one by one.

[0056] When there are too many topological nodes downstream, the verification interval is determined according to the number of topological nodes. For example, if 100 topological nodes are divided into 5 equal parts, the verification interval is 20, which means that each node group has 20 topological nodes. After that, the target node is determined from the node group, and the other topological nodes are simplified to the load of the target node (equivalent to the topological link). The target node is subjected to current balance verification, and it can be quickly determined whether there is a fault line in the node group; if so, the current balance verification can be performed one by one on the topological nodes in the node group; otherwise, the current balance verification of the next node group is performed directly. This embodiment can improve the efficiency of fault location when there are many topological nodes corresponding to the distribution link.

[0057] When there are many topological nodes and the target node is determined and the fault line is located through the node group, when the current balance of the target node is verified, the current is collected through the intelligent switch on the topological link in the same node group as much as possible. When necessary, the current loss in the distribution line can be considered. It should be noted that when locating the fault line through the node group, there is no need to set too many intelligent switches in the distribution line. It is sufficient to ensure that there are available intelligent switches before and after the node group in the distribution link. A small number of intelligent switches are used to determine whether there is a fault in the distribution link, reduce data interaction between intelligent switches, and further locate the fault through the node group and improve the efficiency of fault location.

[0058] In a preferred embodiment, feature extraction is performed on electrical data before and after the line fault in the recorded data, and integrated to generate a fault feature sequence, including: extracting electrical features of the electrical data before and after the line fault, and the difference in electrical features of the before and after electrical data from the recorded data; integrating the electrical features before and after the line fault and the difference in electrical features to generate a fault feature sequence.

[0059] The electrical data of one cycle (or band) before and after the line fault is extracted from the recorded data. The current waveform, phase, effective value, etc. are extracted from the electrical data of the previous cycle, and the same data is extracted from the electrical data of the next cycle. The characteristic data is digitized and integrated into a fault characteristic sequence. Taking the current waveform as an example, the current waveform of the previous cycle is 1 (indicating that it is consistent with the standard waveform), the current waveform of the next cycle is 0 (indicating that it is inconsistent with the standard waveform), and the waveform difference is 0 (indicating that the waveforms of the previous cycle and the next cycle are not both 1).

[0060] After that, the fault feature sequence can be input into the fault judgment model. The fault judgment model is established based on the artificial intelligence model, including: obtaining standard training data; training the constructed artificial intelligence model with the standard training data, and marking it as a fault judgment model after the training is completed.

[0061] The standard training data includes standard input data consistent with the content attributes of the fault feature sequence, and standard output data representing the fault type; the standard training data is obtained by integrating and extracting historical experience, or by laboratory simulation.

[0062] Working principle of the present invention:

[0063] A power grid topology model is established; distribution lines between adjacent substations in the power grid topology model are marked as distribution links, and each power consumption branch in the distribution link is marked as a topology link.

[0064] Collect the current flowing through the smart switch on the distribution link; when there is an abnormal current change, analyze the current balance of the adjacent topological nodes downstream of the smart switch in real time to determine whether there is a line fault; if yes, upload the recorded waveform data of the faulty line at the same time.

[0065] The recorded data is extracted, and the features of the electrical data before and after the line fault in the recorded data are extracted. The fault feature sequence is integrated to generate the fault feature sequence, and the fault type is determined by combining the fault judgment model.

[0066] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. 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 method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A fault analysis and location system for a distribution network, comprising a central control module and an intelligent acquisition module connected thereto; characterized in that: The central control module establishes a power grid topology model; marks the distribution lines between adjacent substations in the power grid topology model as distribution links, and marks each power consumption branch in the distribution link as a topology link; wherein an intelligent switch is provided on the distribution branch or the power consumption branch, and the substation unit includes a transformer or a substation; The intelligent acquisition module collects the current flowing through the intelligent switch on the distribution link; when the current changes abnormally, it analyzes the current balance of the adjacent topological nodes downstream of the intelligent switch in real time to determine whether there is a line fault; if yes, it uploads the recorded waveform data of the faulty line at the same time; the topological node is the intersection of the distribution link and the topological link; The central control module extracts the recorded data, extracts features from the electrical data before and after the line fault in the recorded data, integrates and generates a fault feature sequence, and determines the fault type in combination with the fault judgment model; the fault judgment model is established based on an artificial intelligence model, and the electrical data includes voltage and current; The real-time analysis of the current balance of the adjacent topological nodes downstream of the intelligent switch to determine whether there is a line fault includes: When there is an abnormal change in the current, several topological nodes corresponding to the downstream of the intelligent switch are identified; wherein the upstream and downstream of the intelligent switch are determined by the current flow direction; Based on Kirchhoff's first law, verify whether the current of the downstream topological node is balanced at the same time, and then determine the fault line according to the current balance verification result; The method of verifying whether the current of the downstream topological node is balanced at the same time based on Kirchhoff's first law includes: Identify the number of downstream topological nodes based on the power grid topology model, and then determine the verification interval; The downstream topological nodes are continuously divided into several node groups according to the verification interval; the most downstream topological node in each node group is used as the target node, and the remaining topological nodes are converted into topological links of the target node; Kirchhoff's first law is used to verify whether the current of the target node is balanced at the same time; if yes, the next target node is verified; if not, the topological nodes in the node group corresponding to the target node are verified one by one.

2. A distribution network fault analysis and location system according to claim 1, characterized in that: The central control module communicates and / or is electrically connected with the intelligent acquisition module; and the intelligent acquisition module is associated with the intelligent switch and configured in a topological link of the distribution network; The intelligent acquisition module collects electrical data through an associated intelligent switch; and the intelligent switch is disconnected when a line fault occurs in the corresponding topological link and the fault current is greater than a current threshold.

3. A distribution network fault analysis and location system according to claim 1, characterized in that: The central control module establishes a power grid topology model, including: Obtaining distribution network laying data through smart terminals; wherein the smart terminals include mobile phones or computers; The distribution network laying data is used to identify the substation units and the corresponding distribution links, as well as the topological nodes and topological links in the distribution links. After integration, they are visualized and a power grid topology model is generated.

4. A distribution network fault analysis and location system according to claim 1, characterized in that: The intelligent acquisition module collects the current flowing through the intelligent switch on the power distribution link and determines whether there is an abnormal change in the current, including: Collect the current flowing through the smart switch in real time; Compare the characteristic data of the current with the corresponding standard characteristics to see if they are consistent; if yes, determine that no abnormality has occurred; if no, determine that current abnormality has occurred; wherein the characteristic data includes waveform or effective value.

5. A distribution network fault analysis and location system according to claim 1, characterized in that: The feature extraction of electrical data before and after the line fault in the recorded data and integration to generate a fault feature sequence includes: Extracting electrical characteristics of electrical data before and after the line fault, and the difference in electrical characteristics of the electrical data before and after the fault from the recorded data; wherein the electrical characteristics include waveform, phase or effective value; The electrical characteristics before and after the line fault and the differences in electrical characteristics are integrated to generate a fault feature sequence.

6. A distribution network fault analysis and location system according to claim 5, characterized in that: The fault judgment model is established based on an artificial intelligence model, including: Acquire standard training data; wherein the standard training data includes standard input data consistent with the content attributes of the fault feature sequence, and standard output data indicating the fault type; The artificial intelligence model constructed by training with standard training data is marked as a fault judgment model after training is completed; wherein the artificial intelligence model includes a BP neural network model or a RBF neural network model.

Images