Fault detection method and device for power distribution network, terminal equipment and storage medium

Abstract

The invention provides a fault detection method and device for a power distribution network, terminal equipment and a storage medium, and belongs to the field of power systems.According to the method, a fault detection decision tree used for power distribution network fault detection and positioning is constructed, and when the power distribution network breaks down, the fault detection decision tree can be used for fault detection and positioning according to operation data and environment data when the power distribution network breaks down; detecting a current fault type, a fault reason and a fault position of the power distribution network, specifically, respectively through a fault classification positioning sub-tree for determining the fault position and the fault type according to the operation data and a fault reason tracing sub-tree for determining the fault reason according to the environment data, the fault position and the fault type of the power distribution network; and the fault classification positioning sub-tree and the fault reason tracing sub-tree are integrated and pruned to generate the fault detection decision-making tree. Therefore, the problems that current power distribution network fault detection is not timely and positioning is wrong can be solved.

Description

Technical Field

[0001] The present invention relates to the field of power systems, and in particular to a fault detection method, device, terminal equipment and storage medium for a distribution network. Background Art

[0002] With the continuous expansion of the scale of power systems and the improvement of their intelligence levels, the distribution network, as a key link in power transmission and distribution, has an important significance for ensuring social and economic development and the daily lives of residents by ensuring its stability and reliability.

[0003] Fault location and fault type identification in distribution networks play a crucial role in repairing grid faults and are crucial for the safe, stable, and efficient operation of distribution networks. However, the frequency of distribution network faults and the complexity and diversity of their causes pose significant challenges to fault detection and location. Traditional fault detection methods often rely on manual experience and simple threshold judgments, resulting in delayed fault detection and incorrect location, making them unable to meet the requirements of modern distribution networks for accurate and efficient fault detection. Summary of the Invention

[0004] The present invention provides a distribution network fault detection method, apparatus, terminal equipment and storage medium, which can solve the current problems of untimely distribution network fault detection and incorrect positioning.

[0005] An embodiment of the present invention discloses a method for detecting a fault in a distribution network, comprising:

[0006] Obtaining operational and environmental data of the distribution network;

[0007] Using a preset fault detection decision tree to make predictions based on the operating data and the environmental data, so that the fault detection decision tree detects the current fault type, fault cause, and fault location of the distribution network based on the operating data and the environmental data;

[0008] The construction of the fault detection decision tree includes:

[0009] Acquire several fault samples of the distribution network; wherein each of the fault samples includes: historical operation data and historical environmental data collected when the distribution network has previously experienced a fault, as well as the historical fault type, historical fault cause, and historical fault location determined after fault detection;

[0010] Constructing a fault classification and location subtree for determining the fault location and the fault type based on the historical operation data, the historical fault location, and the historical fault type in the fault sample;

[0011] Constructing a fault cause tracing subtree for determining the cause of the fault based on the environmental data, fault location, and fault type of the distribution network according to the historical environmental data, the historical fault location, and the historical fault cause in the fault sample;

[0012] The fault classification and location subtree and the fault cause tracing subtree are integrated and pruned to generate the fault detection decision tree.

[0013] Furthermore, the historical operation data includes: historical electrical quantity data, historical operating status of the switch device and switch position;

[0014] The constructing, based on the historical operating data, historical fault locations, and historical fault types in the fault sample, a fault classification and location subtree for determining the fault location and fault type based on the operating data includes:

[0015] generating a topological action feature of each historical action state and the corresponding switch position according to the historical action state and the switch position;

[0016] Generating time domain features, frequency domain features, and transient features of each historical electrical quantity data according to the historical electrical quantity data;

[0017] generating a location code for each historical fault location according to the historical fault location;

[0018] The fault location subtree is constructed by using the position codes of several fault samples and the corresponding topological action features;

[0019] The fault classification subtree is constructed by using the historical fault types, location codes of historical fault locations, time domain features, frequency domain features, and transient features of historical electrical quantity data of several fault samples.

[0020] The fault localization subtree and the fault classification subtree are integrated and pruned to generate the fault classification localization subtree.

[0021] Furthermore, the method uses the historical fault types of several fault samples, the location codes of the historical fault locations, the time domain features, the frequency domain features, and the transient features of the historical electrical quantity data to construct a fault classification subtree, and the construction of the fault classification subtree includes:

[0022] Classifying the fault samples according to the historical fault types to generate a fault type subset corresponding to each fault type;

[0023] Using the position code of the historical fault location, the time domain features, the frequency domain features, and the transient features of the historical electrical quantity data as node splitting features, and repeatedly performing a classification decision tree construction operation based on the node splitting features and the fault type subset until the fault classification subtree is generated;

[0024] Furthermore, the classification decision tree construction operation includes:

[0025] Obtaining a plurality of candidate features and a subset of the fault types to be evaluated; wherein, initially, all node splitting features are candidate features, and all fault types are fault types to be evaluated;

[0026] Calculating the information entropy of all subsets of fault types to be evaluated based on the subset of fault types to be evaluated;

[0027] Calculate the information gain rate of each candidate feature according to the information entropy;

[0028] According to the first feature corresponding to the highest information gain rate, several decision nodes of the current level are constructed, and several candidate features other than the first feature are used as the second features of the current level;

[0029] Determining whether there is a first decision node corresponding to multiple fault types among the plurality of decision nodes;

[0030] If so, select the first fault types corresponding to the first decision node on the far left as the fault types to be evaluated required for the next round of classification decision tree construction operation, and use the second features of the current level as candidate features required for the next round of classification decision tree construction operation;

[0031] If not, the decision nodes are traced back layer by layer. When it is determined that there is a second decision node corresponding to multiple fault types, the second fault type corresponding to the second decision node is used as the fault type to be evaluated required for the next round of classification decision tree construction operation, and the second features of the level where the second decision node is located are used as candidate features required for the next round of classification decision tree construction operation;

[0032] When it is determined that there is no second decision node corresponding to the multiple fault types, the fault classification subtree is generated.

[0033] Furthermore, the integrating and pruning the fault classification and location subtree and the fault cause tracing subtree to generate the fault detection decision tree includes:

[0034] Integrate the fault classification and location subtree and the fault cause tracing subtree to generate an initial fault detection decision tree;

[0035] Based on the plurality of fault samples, an association rule algorithm is used to calculate association rules between historical operation data, historical environmental data, historical fault types, historical fault causes, and historical fault locations;

[0036] Calculating the confidence of each node in the initial fault detection decision tree according to the association rules;

[0037] When it is determined that there are target nodes with confidence levels less than a preset threshold in the initial fault detection decision tree, several of the target nodes are pruned from the initial fault detection decision tree to generate the fault detection decision tree.

[0038] Another embodiment of the present invention further provides a fault detection device for a power distribution network, comprising:

[0039] Data acquisition module, used to obtain operation data and environmental data of the distribution network;

[0040] a fault detection module, configured to input the operating data and the environmental data into a preset fault detection decision tree, so that the fault detection decision tree detects the current fault type, fault cause, and fault location of the distribution network based on the operating data and the environmental data;

[0041] The construction of the fault detection decision tree includes:

[0042] Acquire several fault samples of the distribution network; wherein each of the fault samples includes: historical operation data and historical environmental data collected when the distribution network has previously experienced a fault, as well as the historical fault type, historical fault cause, and historical fault location determined after fault detection;

[0043] Constructing a fault classification and location subtree for determining the fault location and the fault type based on the historical operation data, the historical fault location, and the historical fault type in the fault sample;

[0044] Constructing a fault cause tracing subtree for determining the cause of the fault based on the environmental data, fault location, and fault type of the distribution network according to the historical environmental data, the historical fault location, and the historical fault cause in the fault sample;

[0045] The fault classification and location subtree and the fault cause tracing subtree are integrated and pruned to generate the fault detection decision tree.

[0046] Furthermore, the historical operation data includes: historical electrical quantity data, historical operating status of the switch device and switch position;

[0047] The constructing, based on the historical operating data, historical fault locations, and historical fault types in the fault sample, a fault classification and location subtree for determining the fault location and fault type based on the operating data includes:

[0048] generating a topological action feature of each historical action state and the corresponding switch position according to the historical action state and the switch position;

[0049] Generating time domain features, frequency domain features, and transient features of each historical electrical quantity data according to the historical electrical quantity data;

[0050] generating a location code for each historical fault location according to the historical fault location;

[0051] The fault location subtree is constructed by using the position codes of several fault samples and the corresponding topological action features;

[0052] The fault classification subtree is constructed by using the historical fault types, location codes of historical fault locations, time domain features, frequency domain features, and transient features of historical electrical quantity data of several fault samples.

[0053] The fault localization subtree and the fault classification subtree are integrated and pruned to generate the fault classification localization subtree.

[0054] Furthermore, the method uses the historical fault types of several fault samples, the location codes of the historical fault locations, the time domain features, the frequency domain features, and the transient features of the historical electrical quantity data to construct a fault classification subtree, and the construction of the fault classification subtree includes:

[0055] Classifying the fault samples according to the historical fault types to generate a fault type subset corresponding to each fault type;

[0056] The position code of the historical fault location, the time domain features, the frequency domain features, and the transient features of the historical electrical quantity data are used as node splitting features, and the classification decision tree construction operation is repeated according to the node splitting features and the fault type subset until the fault classification subtree is generated.

[0057] Another embodiment of the present invention further provides a terminal device, comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps of a distribution network fault detection method as described in the above embodiment of the present invention.

[0058] Another embodiment of the present invention further provides a computer-readable storage medium item, comprising: a stored computer program, which, when the computer program is running, controls the device where the computer-readable storage medium is located to execute the steps of a distribution network fault detection method as described in the above embodiment of the present invention.

[0059] The following beneficial effects are achieved by implementing the present invention:

[0060] The present invention provides a fault detection method, apparatus, terminal device, and storage medium for a distribution network. The method constructs a fault detection decision tree for distribution network fault detection and location. When a distribution network fault occurs, the method can detect the current fault type, fault cause, and fault location of the distribution network based on the operating data and environmental data at the time of the distribution network fault. Specifically, the method comprises a fault classification and location subtree for determining the fault location and fault type based on the operating data, and a fault cause tracing subtree for determining the fault cause based on the environmental data, fault location, and fault type of the distribution network. The fault classification and location subtree and the fault cause tracing subtree are integrated and pruned to generate the fault detection decision tree. Therefore, the fault detection decision tree disclosed by the present invention can locate the fault location and determine the fault type of the distribution network based on the operating data collected when the distribution network fails. The predicted fault location and fault type are then combined with the environmental data to predict the fault cause. The method can timely detect distribution network faults and provide the fault cause, thereby improving the efficiency of fault repair and effectively improving the safety and stability of distribution network operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the technical solution of the present application, the following is a brief introduction to the drawings required for use in the implementation. Obviously, the drawings described below are only some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0062] Figure 1 This is a flow chart of a method for detecting a fault in a distribution network provided by one embodiment of the present invention;

[0063] Figure 2 This is a structural diagram of a fault detection device for a distribution network provided by one embodiment of the present invention;

[0064] Figure 3 This is a schematic diagram of a distribution network topology structure provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0065] To make the objectives, technical solutions, and advantages of this application more clear, the technical solutions in this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments 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 any creative efforts are within the scope of protection of this application.

[0066] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned figure descriptions are intended to cover non-exclusive inclusions.

[0067] In the description of the embodiments of this application, the technical terms "first" and "second" are used only to distinguish different objects and should not be understood to indicate or imply relative importance or implicitly specify the quantity, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, the meaning of "plurality" is more than two, unless otherwise clearly and specifically defined.

[0068] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0069] In the description of the embodiments of this application, the term "and / or" is simply a description of the association relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent the following three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally indicates that the associated objects are in an "or" relationship.

[0070] In the description of the embodiments of the present application, the term "multiple" refers to more than two (including two). Similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).

[0071] In the description of the embodiments of the present application, unless otherwise expressly specified or limited, technical terms such as "installed," "connected," "connected," and "fixed" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integration; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; internal connections between two components or interactions between two components. Those skilled in the art can understand the specific meanings of the above terms in the embodiments of the present application based on specific circumstances.

[0072] See also Figure 1 To solve the current problems of delayed fault detection and incorrect positioning in distribution networks, an embodiment of the present invention provides a method for fault detection in distribution networks, comprising:

[0073] S1. Obtaining the operation data and environmental data of the distribution network;

[0074] In a preferred embodiment of the present invention, the operating data includes: electrical quantity data, the operating status of the switching device, and the switch position. It is understood that the electrical quantity data includes the power flow of each transmission line, the output current and voltage of each distribution node, and the output current and voltage, etc., and the switching device is an automatic switching device for disconnecting the fault line, such as a circuit breaker. It should be noted that in order to reduce the difficulty of training the subsequent decision tree, each node is numbered and each transmission line is coded according to the node to which it is connected. It is understood that, if Figure 3 As shown, the line connecting node 1 and node 2 is coded as 1-2, and then each switching line is coded according to the line where each switching device is located, and the switching position of the switching device is determined by the code.

[0075] Furthermore, environmental data includes weather, humidity, light intensity, etc. It has been verified that distribution network faults are highly correlated with weather. To ensure the operating efficiency of the fault detection decision tree, in this embodiment, only weather is used as the input feature of the fault detection decision tree. Similarly, to reduce the difficulty of training subsequent decision trees, the English abbreviations corresponding to the weather types are used to encode various types of weather.

[0076] It should be noted that in order to ensure the accuracy of subsequent fault diagnosis, the operating data and environmental data are preprocessed before feature extraction, including missing value detection and filling, and outlier detection and cleaning. Specifically, data processing tools are used to detect whether there are null values ​​or invalid values ​​in the data set. If null values ​​or invalid values ​​are detected, they are deleted, and support vector machines are used to predict missing data. Use existing data for training, take data containing missing values ​​as input, and predict and supplement missing data values. First, standardize data of different dimensions, then normalize the standardized data, and finally detect the timestamp and numerical format of the data format to ensure data cleaning and format uniformity.

[0077] S2. Using a preset fault detection decision tree to perform prediction based on the operating data and the environmental data, so that the fault detection decision tree detects the current fault type, fault cause, and fault location of the distribution network based on the operating data and the environmental data;

[0078] In a preferred embodiment of the present invention, feature extraction is performed on the operating data. Specifically, the time domain features (such as the peak value, effective value, and waveform distortion rate of the current / voltage), frequency domain features (such as the harmonic components analyzed by FFT), and transient features (such as the zero-sequence current mutation rate and the voltage phase angle offset) of the electrical quantity data in the operating data are extracted. According to the operating state and switch position of the switching device, the topological action features are constructed. For example, the action timing sequence of the switching device is constructed according to the position and action sequence of the switch. Finally, the environmental data is encoded and the input matrix of the fault detection decision tree is constructed together with the time domain features, frequency domain features, and transient features of the electrical quantity data and the topological action features of the switching device, so that the fault detection decision tree can detect the current fault type, fault cause, and fault location of the distribution network according to the input matrix.

[0079] The construction of the fault detection decision tree includes:

[0080] S01. Acquire several fault samples of the distribution network; wherein each of the fault samples includes: historical operation data and historical environmental data collected when a fault occurred in the distribution network in the past, as well as the historical fault type, historical fault cause, and historical fault location determined after fault detection;

[0081] In a preferred embodiment of the present invention, in order to ensure the efficiency and accuracy of constructing the fault detection decision tree, the same method as step S1 is used to perform data preprocessing on the fault samples.

[0082] It should be noted that this embodiment defines each fault type in detail according to its occurrence mechanism, impact range, repair difficulty and other characteristics, so as to facilitate subsequent feature extraction and decision tree construction. Therefore, the set fault types include: short circuit fault, ground fault, equipment damage, etc.; further, this embodiment selects fault samples according to the cause of the fault. Specifically, based on the failure factors such as the failure rate of the equipment and load constraints, an equipment factor data set of equipment-related faults is constructed; considering the impact of the environment on distribution network failures, the susceptibility of equipment failures under different environmental conditions is analyzed to construct an environmental factor data set; analyzing factors such as load fluctuations and the operating pressure of the distribution network, a network load factor data set is constructed; analyzing human factors such as operational errors, inadequate maintenance, and improper fault handling, a human factor data set is constructed.

[0083] S02. Constructing a fault classification and location subtree for determining the fault location and fault type based on the historical operation data, historical fault location, and historical fault type in the fault sample;

[0084] Preferably, the historical operation data includes: historical electrical quantity data, historical operation status of the switch device and switch position;

[0085] The constructing, based on the historical operating data, historical fault locations, and historical fault types in the fault sample, a fault classification and location subtree for determining the fault location and fault type based on the operating data includes:

[0086] S021. Generate a topological action feature of each historical action state and corresponding switch position according to the historical action state and the switch position;

[0087] S022. Generate time domain characteristics, frequency domain characteristics, and transient characteristics of each historical electrical quantity data based on the historical electrical quantity data;

[0088] S023. Generate a location code for each historical fault location based on the historical fault location;

[0089] S024. Construct a fault location subtree using the position codes of several fault samples and corresponding topological action features;

[0090] S025. Construct a fault classification subtree using historical fault types, location codes of historical fault locations, time domain features, frequency domain features, and transient features of historical electrical quantity data of several fault samples;

[0091] In a preferred embodiment of the present invention, a switch device in a distribution network automatically trips upon detecting a sudden fault on its line, promptly disconnecting the faulty line and preventing the fault from spreading to other distribution areas. Therefore, the location of the faulty line or node can be determined based on the switch device's position and sequence of operation. Therefore, this embodiment constructs a fault location subtree based on historical action states and corresponding topological action characteristics of switch positions, enabling accurate location of the distribution network fault.

[0092] Preferably, the method of constructing a fault classification subtree by using historical fault types, location codes of historical fault locations, time domain features, frequency domain features, and transient features of historical electrical quantity data of several fault samples includes:

[0093] S0251. Classify the fault samples according to the historical fault types, and generate a fault type subset corresponding to each fault type;

[0094] S0252. Using the location code of the historical fault location, the time domain features, the frequency domain features, and the transient features of the historical electrical quantity data as node splitting features, and repeatedly performing a classification decision tree construction operation based on the node splitting features and the fault type subset until the fault classification subtree is generated;

[0095] Preferably, the classification decision tree construction operation includes:

[0096] S02521. Obtain several candidate features and a subset of several fault types to be evaluated; wherein, initially, all node splitting features are candidate features, and all fault types are fault types to be evaluated;

[0097] S02522. Calculate the information entropy of all subsets of fault types to be evaluated based on the subset of fault types to be evaluated;

[0098] S02523. Calculate the information gain rate of each candidate feature based on the information entropy;

[0099] S02524. Construct several decision nodes at the current level based on the first feature corresponding to the highest information gain rate, and use several candidate features other than the first feature as the second features at the current level;

[0100] S02525. Determine whether there is a first decision node corresponding to multiple fault types among the plurality of decision nodes;

[0101] S02526. If yes, select the first fault types corresponding to the first decision node on the far left as the fault types to be evaluated required for the next round of classification decision tree construction operation, and use the second features of the current level as candidate features required for the next round of classification decision tree construction operation;

[0102] If not, then backtrack the decision nodes layer by layer. When it is determined that there is a second decision node corresponding to multiple fault types, the second fault type corresponding to the second decision node is used as the fault type to be evaluated required for the next round of classification decision tree construction operation, and the plurality of second features at the level where the second decision node is located are used as candidate features required for the next round of classification decision tree construction operation;

[0103] S02528. When it is determined that there is no second decision node corresponding to multiple types of faults, generate the fault classification subtree.

[0104] In a preferred embodiment of the present invention, the information entropy of all subsets of fault types to be evaluated is calculated according to the following formula; the smaller the information entropy, the purer the data set:

[0105]

[0106] Where: E is the information entropy, N is the total number of all fault type subsets to be evaluated, P n is the probability of occurrence of samples of the nth subset of fault types to be evaluated.

[0107] Furthermore, the information gain rate of each candidate feature is calculated according to the following formula:

[0108]

[0109] in:

[0110]

[0111] Where: GR is the information gain rate, S is the amount of information, G is the information gain, D is the data set consisting of all subsets of fault types to be evaluated, D n is the nth subset of fault types to be evaluated in the dataset.

[0112] S026: Integrate and prune the fault localization subtree and the fault classification subtree to generate the fault classification localization subtree.

[0113] In a preferred embodiment of the present invention, the same method as steps S041 to S044 is used to integrate and prune the fault localization subtree and the fault classification subtree to generate the fault classification localization subtree.

[0114] S03. Constructing a fault cause tracing subtree for determining the cause of the fault based on the environmental data, fault location, and fault type of the distribution network according to the historical environmental data, fault location, and fault type in the fault sample, and the historical fault cause;

[0115] In a preferred embodiment of the present invention, historical environmental data, historical fault locations, and historical fault types are encoded to construct candidate features of the fault cause tracing subtree, and then the fault cause tracing subtree is recursively constructed using the same method as steps S02521-S02528.

[0116] S04: Integrate and prune the fault classification and location subtree and the fault cause tracing subtree to generate the fault detection decision tree.

[0117] Preferably, integrating and pruning the fault classification and location subtree and the fault cause tracing subtree to generate the fault detection decision tree includes:

[0118] S041. Integrate the fault classification and location subtree and the fault cause tracing subtree to generate an initial fault detection decision tree;

[0119] S042. Calculate association rules among historical operating data, historical environmental data, historical fault types, historical fault causes, and historical fault locations using an association rule algorithm based on the plurality of fault samples;

[0120] S043. Calculate the confidence of each node in the initial fault detection decision tree according to the association rule;

[0121] S044. When it is determined that there are target nodes with confidence levels less than a preset threshold in the initial fault detection decision tree, prune several of the target nodes from the initial fault detection decision tree to generate the fault detection decision tree.

[0122] In a preferred embodiment of the present invention, based on a number of the fault samples, an association rule algorithm is used to calculate association rules between historical operating data, historical environmental data, historical fault types, historical fault causes, and historical fault locations. The specific steps are as follows:

[0123] Association rules are shaped like A→B, which reflects the regularity of the occurrence of elements in B when elements in A appear. And there is a dependency relationship between A and B, then define A→B as an association rule. The association rule construction method of this method is as follows:

[0124] (1) Calculate support: The support X is expressed as the probability that A→B is included in C. The calculation method is as follows:

[0125]

[0126] Where count(A∪B) is the number of times A or B appears in C, and C is the total number of records in the transaction database.

[0127] (2) Calculate the confidence level: The confidence level Y is the ratio of the number of cases in C that contain A→B to the number of cases that contain A. The calculation method is as follows:

[0128]

[0129] Where: P(A∩B) represents the probability that A and B appear in C at the same time; P(A) represents the probability that A appears in C.

[0130] (3) Calculate the lift: The lift L is the ratio of the confidence that A→B is included in C to the number of B included in C. The calculation method is as follows:

[0131]

[0132] Where: P(B) is the probability that B appears in C.

[0133] Furthermore, the aforementioned association rule algorithm is used to scan the fault samples and generate frequent item sets. These item sets represent the frequent joint occurrence relationships between different features. Finally, association rules are extracted from the frequent item sets. These rules reflect the relationship between the fault type and its possible cause.

[0134] Finally, the effectiveness of each decision node is evaluated using an association rule algorithm. If the confidence level of the association rule for a decision node is low, the node is considered to contribute little to the classification of the decision tree and can be pruned. If the rule corresponding to a node in a branch has low support or confidence after association rule analysis, the node and its subtree can be deleted to reduce the complexity of the tree.

[0135] Furthermore, this embodiment uses the historical operation and maintenance data of a region's distribution network equipment module fault data to conduct case analysis, and the distribution network line fitting constructs the IEEE33 multi-terminal distribution network structure, such as Figure 3 As shown, numbers 1-33 are line nodes. The system includes multiple source systems such as wind power generation, photovoltaic power generation, energy storage system, and electric vehicle charging station. The method proposed in this invention is verified by simulating the normal state and fault state of the system. First, a phase-to-phase short circuit fault is simulated on the detection line. The method of this invention and the CNN artificial intelligence algorithm are compared and analyzed. The results are shown in the following table:

[0136]

[0137] It can be seen from the table that the method proposed in the present invention is superior to the CNN artificial intelligence algorithm in terms of fault location and fault identification success rate, which can make the distribution network safer and more stable.

[0138] like Figure 2 As shown, based on the above method embodiment, a corresponding device embodiment is provided;

[0139] An embodiment of the present invention provides a fault detection device for a power distribution network, comprising:

[0140] Data acquisition module, used to obtain operation data and environmental data of the distribution network;

[0141] a fault detection module, configured to input the operating data and the environmental data into a preset fault detection decision tree, so that the fault detection decision tree detects the current fault type, fault cause, and fault location of the distribution network based on the operating data and the environmental data;

[0142] The construction of the fault detection decision tree includes:

[0143] Acquire several fault samples of the distribution network; wherein each of the fault samples includes: historical operation data and historical environmental data collected when the distribution network has previously experienced a fault, as well as the historical fault type, historical fault cause, and historical fault location determined after fault detection;

[0144] Constructing a fault classification and location subtree for determining the fault location and the fault type based on the historical operation data, the historical fault location, and the historical fault type in the fault sample;

[0145] Constructing a fault cause tracing subtree for determining the cause of the fault based on the environmental data, fault location, and fault type of the distribution network according to the historical environmental data, the historical fault location, and the historical fault cause in the fault sample;

[0146] The fault classification and location subtree and the fault cause tracing subtree are integrated and pruned to generate the fault detection decision tree.

[0147] Preferably, the historical operation data includes: historical electrical quantity data, historical operation status of the switch device and switch position;

[0148] The constructing, based on the historical operating data, historical fault locations, and historical fault types in the fault sample, a fault classification and location subtree for determining the fault location and fault type based on the operating data includes:

[0149] generating a topological action feature of each historical action state and the corresponding switch position according to the historical action state and the switch position;

[0150] Generating time domain features, frequency domain features, and transient features of each historical electrical quantity data according to the historical electrical quantity data;

[0151] generating a location code for each historical fault location according to the historical fault location;

[0152] The fault location subtree is constructed by using the position codes of several fault samples and the corresponding topological action features;

[0153] The fault classification subtree is constructed by using the historical fault types, location codes of historical fault locations, time domain features, frequency domain features, and transient features of historical electrical quantity data of several fault samples.

[0154] The fault localization subtree and the fault classification subtree are integrated and pruned to generate the fault classification localization subtree.

[0155] Preferably, the method of constructing a fault classification subtree by using historical fault types, location codes of historical fault locations, time domain features, frequency domain features, and transient features of historical electrical quantity data of several fault samples includes:

[0156] Classifying the fault samples according to the historical fault types to generate a fault type subset corresponding to each fault type;

[0157] The position code of the historical fault location, the time domain features, the frequency domain features, and the transient features of the historical electrical quantity data are used as node splitting features, and the classification decision tree construction operation is repeated according to the node splitting features and the fault type subset until the fault classification subtree is generated.

[0158] It can be understood that the above-mentioned device embodiment corresponds to the method embodiment of the present invention, and can implement any one of the above-mentioned method embodiments of the present invention to provide a distribution network fault detection method.

[0159] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. Furthermore, in the drawings of the device embodiments provided by the present invention, the connection relationship between modules indicates that they have a communication connection, which may be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement the present invention without inventive effort.

[0160] Based on the above-mentioned embodiment of a distribution network fault detection method, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, a distribution network fault detection method of any embodiment of the present invention is implemented.

[0161] For example, in this embodiment, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to implement the present invention. The one or more module elements may be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program in the terminal device.

[0162] The terminal device may be a computing device such as a desktop computer, a notebook computer, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0163] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, connecting various parts of the entire terminal device using various interfaces and lines.

[0164] Based on the above method embodiments, another embodiment of the present invention provides a computer-readable storage medium, including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute a distribution network fault detection method described in any one of the above method embodiments of the present invention.

[0165] Wherein, the module / unit integrated in the device / terminal equipment, if implemented in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, it can implement the steps of the above-mentioned various method embodiments. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0166] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for detecting a fault in a distribution network, characterized in that: include: Obtaining operational and environmental data of the distribution network; Using a preset fault detection decision tree to make predictions based on the operating data and the environmental data, so that the fault detection decision tree detects the current fault type, fault cause, and fault location of the distribution network based on the operating data and the environmental data; The construction of the fault detection decision tree includes: Acquire several fault samples of the distribution network; wherein each of the fault samples includes: historical operation data and historical environmental data collected when the distribution network has previously experienced a fault, as well as the historical fault type, historical fault cause, and historical fault location determined after fault detection; Constructing a fault classification and location subtree for determining the fault location and the fault type based on the historical operation data, the historical fault location, and the historical fault type in the fault sample; Constructing a fault cause tracing subtree for determining the cause of the fault based on the environmental data, fault location, and fault type of the distribution network according to the historical environmental data, the historical fault location, and the historical fault cause in the fault sample; The fault classification and location subtree and the fault cause tracing subtree are integrated and pruned to generate the fault detection decision tree.

2. A method for detecting a fault in a distribution network according to claim 1, characterized in that: The historical operation data includes: historical electrical quantity data, historical operation status of the switch device and switch position; The constructing, based on the historical operating data, historical fault locations, and historical fault types in the fault sample, a fault classification and location subtree for determining the fault location and fault type based on the operating data includes: generating a topological action feature of each historical action state and the corresponding switch position according to the historical action state and the switch position; Generating time domain features, frequency domain features, and transient features of each historical electrical quantity data according to the historical electrical quantity data; generating a location code for each historical fault location according to the historical fault location; The fault location subtree is constructed by using the position codes of several fault samples and the corresponding topological action features; The fault classification subtree is constructed by using the historical fault types, location codes of historical fault locations, time domain features, frequency domain features, and transient features of historical electrical quantity data of several fault samples. The fault localization subtree and the fault classification subtree are integrated and pruned to generate the fault classification localization subtree.

3. A method for detecting a fault in a distribution network according to claim 2, characterized in that: The method uses the historical fault types, location codes of historical fault locations, time domain features, frequency domain features, and transient features of historical electrical quantity data of several fault samples to construct a fault classification subtree, including: Classifying the fault samples according to the historical fault types to generate a fault type subset corresponding to each fault type; The position code of the historical fault location, the time domain features, the frequency domain features, and the transient features of the historical electrical quantity data are used as node splitting features, and the classification decision tree construction operation is repeated according to the node splitting features and the fault type subset until the fault classification subtree is generated.

4. A method for detecting a fault in a distribution network according to claim 3, characterized in that: The classification decision tree construction operation includes: Obtaining a plurality of candidate features and a subset of the fault types to be evaluated; wherein, initially, all node splitting features are candidate features, and all fault types are fault types to be evaluated; Calculating the information entropy of all subsets of fault types to be evaluated based on the subset of fault types to be evaluated; Calculate the information gain rate of each candidate feature according to the information entropy; According to the first feature corresponding to the highest information gain rate, several decision nodes of the current level are constructed, and several candidate features other than the first feature are used as the second features of the current level; Determining whether there is a first decision node corresponding to multiple fault types among the plurality of decision nodes; If so, select the first fault types corresponding to the first decision node on the far left as the fault types to be evaluated required for the next round of classification decision tree construction operation, and use the second features of the current level as candidate features required for the next round of classification decision tree construction operation; If not, the decision nodes are traced back layer by layer. When it is determined that there is a second decision node corresponding to multiple fault types, the second fault type corresponding to the second decision node is used as the fault type to be evaluated required for the next round of classification decision tree construction operation, and the second features of the level where the second decision node is located are used as candidate features required for the next round of classification decision tree construction operation; When it is determined that there is no second decision node corresponding to the multiple fault types, the fault classification subtree is generated.

5. A method for detecting a fault in a distribution network according to claim 4, characterized in that: The integrating and pruning the fault classification and location subtree and the fault cause tracing subtree to generate the fault detection decision tree includes: Integrate the fault classification and location subtree and the fault cause tracing subtree to generate an initial fault detection decision tree; Based on the plurality of fault samples, an association rule algorithm is used to calculate association rules between historical operation data, historical environmental data, historical fault types, historical fault causes, and historical fault locations; Calculating the confidence of each node in the initial fault detection decision tree according to the association rules; When it is determined that there are target nodes with confidence levels less than a preset threshold in the initial fault detection decision tree, several of the target nodes are pruned from the initial fault detection decision tree to generate the fault detection decision tree.

6. A fault detection device for a distribution network, characterized in that: include: Data acquisition module, used to obtain operation data and environmental data of the distribution network; a fault detection module, configured to input the operating data and the environmental data into a preset fault detection decision tree, so that the fault detection decision tree detects the current fault type, fault cause, and fault location of the distribution network based on the operating data and the environmental data; The construction of the fault detection decision tree includes: Acquire several fault samples of the distribution network; wherein each of the fault samples includes: historical operation data and historical environmental data collected when the distribution network has previously experienced a fault, as well as the historical fault type, historical fault cause, and historical fault location determined after fault detection; Constructing a fault classification and location subtree for determining the fault location and the fault type based on the historical operation data, the historical fault location, and the historical fault type in the fault sample; Constructing a fault cause tracing subtree for determining the cause of the fault based on the environmental data, fault location, and fault type of the distribution network according to the historical environmental data, the historical fault location, and the historical fault cause in the fault sample; The fault classification and location subtree and the fault cause tracing subtree are integrated and pruned to generate the fault detection decision tree.

7. A fault detection device for a power distribution network according to claim 6, characterized in that: The historical operation data includes: historical electrical quantity data, historical operation status of the switch device and switch position; The constructing, based on the historical operating data, historical fault locations, and historical fault types in the fault sample, a fault classification and location subtree for determining the fault location and fault type based on the operating data includes: generating a topological action feature of each historical action state and the corresponding switch position according to the historical action state and the switch position; Generating time domain features, frequency domain features, and transient features of each historical electrical quantity data according to the historical electrical quantity data; generating a location code for each historical fault location according to the historical fault location; The fault location subtree is constructed by using the position codes of several fault samples and the corresponding topological action features; The fault classification subtree is constructed by using the historical fault types, location codes of historical fault locations, time domain features, frequency domain features, and transient features of historical electrical quantity data of several fault samples. The fault localization subtree and the fault classification subtree are integrated and pruned to generate the fault classification localization subtree.

8. A fault detection device for a power distribution network as claimed in claim 7, characterized in that: The method uses the historical fault types of several fault samples, the location codes of historical fault locations, the time domain features, the frequency domain features, and the transient features of historical electrical quantity data to construct a fault classification subtree, and the fault classification subtree includes: Classifying the fault samples according to the historical fault types to generate a fault type subset corresponding to each fault type; The position code of the historical fault location, the time domain features, the frequency domain features, and the transient features of the historical electrical quantity data are used as node splitting features, and the classification decision tree construction operation is repeated according to the node splitting features and the fault type subset until the fault classification subtree is generated.

9. A terminal device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method for detecting faults in a distribution network according to any one of claims 1 to 5 is implemented.

10. A computer-readable storage medium, characterized in that include: A stored computer program, wherein when the computer program is run, the device where the computer-readable storage medium is located is controlled to execute a distribution network fault detection method according to any one of claims 1 to 5.

Images