Intelligent fault location method and system based on multi-source power information integration

Abstract

The present invention discloses an intelligent fault location method and system based on multi-source power information integration, which relates to the technical field of power fault location. The method and system collect line parameter data, node load data and equipment status information to form line auxiliary information, collect historical fault data, build a fault prediction Bayesian network model, and use the historical fault data to train the fault prediction Bayesian network model to generate a node conditional probability table, collect real-time monitoring data of power lines, and map the real-time monitoring data to network node evidence. The real-time acquired network node evidence is input into the fault prediction Bayesian network model, and the fault probability of each network node is calculated based on the node conditional probability table, and nodes with high abnormal probability are screened out as fault investigation nodes; the risk of errors in the states provided by network nodes can be reduced, and the accuracy of power fault location can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power fault location, and in particular to an intelligent fault location method and system based on multi-source power information integration. Background Art

[0002] The power system is an important infrastructure for the operation of modern society. Any line failure may lead to large-scale power outages, causing huge economic losses and social impacts. Therefore, real-time monitoring of power lines and rapid and accurate fault location are crucial.

[0003] However, traditional fault location methods mainly rely on manual experience analysis and on-site inspection, which is inefficient and easily affected by human factors. As the scale of power grids continues to expand, line parameters, loads and other information are becoming more and more complex, and single-dimensional analysis can no longer meet the needs of accurate fault location.

[0004] Existing technologies usually only focus on a single information source (such as monitoring data), ignoring the value of other important information such as line parameters and equipment status, resulting in insufficient accuracy in fault cause analysis. At the same time, since the real-time status update of line parameter units such as various power equipment and load terminals may have errors or real-time problems, further errors may occur in fault location.

[0005] The Chinese patent with publication number CN114595839A discloses a method and system for analyzing the cause of faults of smart electric energy meters, including: obtaining fault-related information of the current fault; calculating the similarity between the fault-related information of the current fault and each association rule in the pre-constructed association rule set; determining the fault cause of the current fault based on the similarity; wherein the pre-constructed association rule set is based on the fault causes and fault-related information of the historical faults of the smart electric energy meter, and is constructed by mining, correcting and eliminating redundant rules of association rules between the fault causes and fault-related information. However, this method is based on known association rules and lacks more intelligent association rules for mining fault-related relationships of various power equipment;

[0006] To this end, the present invention proposes an intelligent fault location method and system based on multi-source power information integration. Summary of the invention

[0007] 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 an intelligent fault location method and system based on multi-source power information integration, which can reduce the risk of errors in the status provided by network nodes and improve the accuracy of power fault location.

[0008] To achieve the above object, the present invention proposes an intelligent fault location method based on multi-source power information integration, comprising the following steps:

[0009] Step 1: Collect line parameter data, node load data and equipment status information in advance to form line auxiliary information, and collect historical fault data in advance;

[0010] Step 2: Build a fault prediction Bayesian network model based on line auxiliary information, and use historical fault data to train the fault prediction Bayesian network model to generate a node conditional probability table;

[0011] Step 3: Collect real-time monitoring data of power lines and map the real-time monitoring data into network node evidence;

[0012] Step 4: Input the real-time acquired network node evidence into the fault prediction Bayesian network model, calculate the fault probability of each network node based on the node conditional probability table, and select nodes with high abnormal probability as fault troubleshooting nodes;

[0013] The line parameter data is collected in the following manner:

[0014] Collect the line electrical parameters, line structural parameters and line corridor environmental conditions of the power line to be monitored to form line parameter data;

[0015] The node load data is collected in the following manner:

[0016] Determine important bus nodes and key user nodes in power lines;

[0017] Collect node load curves of important busbar nodes;

[0018] Collect user electricity consumption data of key user nodes;

[0019] Collecting the time attributes of each important load node; the important load nodes include important bus nodes and key user nodes;

[0020] The node load data is composed of the node load curve of the important bus nodes, the user power consumption data of the key user nodes and the time attributes of the important load nodes;

[0021] The device status data is collected in the following manner:

[0022] Collect the defects and fault status of each known power equipment in the power line, equipment maintenance plan and records, and basic equipment information to form equipment status data;

[0023] The historical fault data is collected in the following manner:

[0024] Collect basic information about past power failures in power lines, collect the operating status of each power device when each power failure occurs, collect the failure evolution process and processing records, and mark the cause of the power failure to form historical failure data;

[0025] The construction of the fault prediction Bayesian network model based on line auxiliary information includes the following steps:

[0026] Step 11: Based on the line parameter data and node load data, define the network nodes and their initial states of the fault prediction Bayesian network model;

[0027] Step 12: Based on the device status data, design the node status variables and their value ranges for each network node;

[0028] Step 13: Build the topological dependency relationship between each node;

[0029] The topological dependency relationship between the nodes is constructed in the following manner:

[0030] Using the structural information in the line parameter data, physical connection dependencies are established;

[0031] Using the power supply relationship in the node load data, the load-power supply dependency relationship is established;

[0032] The method of using historical fault data to train the fault prediction Bayesian network model and generate a node conditional probability table comprises the following steps:

[0033] Step 21: Obtain the node status of each node at each power failure from the historical fault case data to obtain the fault state vector at each power failure;

[0034] Step 22: Input each fault state vector into the fault prediction Bayesian network model according to the corresponding relationship of the nodes, perform CPT parameter estimation on the CPT parameters of each node, and train the model using parameter learning and structure optimization algorithms until the CPT parameter estimation value of the node reaches convergence and the training is stopped when the optimal network model structure is generated;

[0035] Step 23: Read the CPT parameters of each node to obtain the final node conditional probability table;

[0036] The method of collecting real-time monitoring data of power lines is:

[0037] Collect real-time node load data and real-time equipment status information of power lines during real-time operation to form real-time monitoring data;

[0038] The method of inputting the network node evidence acquired in real time into the fault prediction Bayesian network model and calculating the fault probability of each network node based on the node conditional probability table is as follows:

[0039] Input the fault location feature vector into the fault prediction Bayesian network model;

[0040] Determine the Markov spiral solution order of each network node according to the network topology of the fault prediction Bayesian network model;

[0041] Traverse each network node in order, query the conditional probability from the node conditional probability table of its parent node as the failure probability of the network node;

[0042] The method of selecting nodes with high abnormal probability as fault troubleshooting nodes is as follows:

[0043] A fault probability threshold is set in advance, and all network nodes whose fault probabilities are greater than the fault probability threshold are selected as fault troubleshooting nodes, and the fault troubleshooting nodes are sorted from large to small according to the fault probability.

[0044] An intelligent fault location system based on multi-source power information integration is proposed, including a basic data collection module, a Bayesian network model training module and an intelligent fault location module; wherein each module is electrically connected;

[0045] The basic data collection module collects line parameter data, node load data and equipment status information in advance to form line auxiliary information, and collects historical fault data in advance and sends the historical fault data to the Bayesian network model training module;

[0046] The Bayesian network model training module builds a fault prediction Bayesian network model based on line auxiliary information, and uses historical fault data to train the fault prediction Bayesian network model, generates a node conditional probability table, and sends the fault prediction Bayesian network model and the node conditional probability table to the intelligent fault location module;

[0047] The intelligent fault location module collects real-time monitoring data of power lines and maps the real-time monitoring data into network node evidence. It inputs the real-time acquired network node evidence into the fault prediction Bayesian network model, calculates the fault probability of each network node based on the node conditional probability table, and selects nodes with high abnormal probability as fault detection nodes.

[0048] The present invention provides an electronic device, comprising: a processor and a memory, wherein the memory stores a computer program that can be called by the processor;

[0049] The processor executes the above-mentioned intelligent fault location method based on multi-source power information integration by calling the computer program stored in the memory.

[0050] The present invention provides a computer-readable storage medium having a rewritable computer program stored thereon;

[0051] When the computer program is executed on a computer device, the computer device is enabled to execute the above-mentioned intelligent fault location method based on multi-source power information integration.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] The present invention collects line parameter data, node load data and equipment status information in advance to form line auxiliary information, collects historical fault data in advance, builds a fault prediction Bayesian network model based on the line auxiliary information, and uses the historical fault data to train the fault prediction Bayesian network model to generate a node conditional probability table, collects real-time monitoring data of the power line, and maps the real-time monitoring data to network node evidence, inputs the real-time acquired network node evidence into the fault prediction Bayesian network model, calculates the failure probability of each network node based on the node conditional probability table, and selects nodes with high abnormal probability as fault investigation nodes; by integrating multi-source information such as line parameters, load and equipment status, and combining historical fault data to train the Bayesian network model, the risk of errors in the status provided by the network nodes can be reduced, and the accuracy of power fault location can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a flow chart of an intelligent fault location method based on multi-source power information integration in Embodiment 1 of the present invention;

[0055] Figure 2 This is a module connection relationship diagram of an intelligent fault location system based on multi-source power information integration in Example 2 of the present invention;

[0056] Figure 3 This is a schematic diagram of the structure of an electronic device in Embodiment 3 of the present invention;

[0057] Figure 4 This is a schematic diagram of the computer-readable storage medium structure in Example 4 of the present invention. DETAILED DESCRIPTION

[0058] 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.

[0059] Example 1

[0060] like Figure 1 As shown, the intelligent fault location method based on multi-source power information integration includes the following steps:

[0061] Step 1: Collect line parameter data, node load data and equipment status information in advance to form line auxiliary information, and collect historical fault data in advance;

[0062] Step 2: Build a fault prediction Bayesian network model based on line auxiliary information, and use historical fault data to train the fault prediction Bayesian network model to generate a node conditional probability table;

[0063] Step 3: Collect real-time monitoring data of power lines and map the real-time monitoring data into network node evidence;

[0064] Step 4: Input the real-time acquired network node evidence into the fault prediction Bayesian network model, calculate the fault probability of each network node based on the node conditional probability table, and select nodes with high abnormal probability as fault troubleshooting nodes;

[0065] The line parameter data is collected in the following manner:

[0066] Collect the line electrical parameters, line structural parameters and line corridor environmental conditions of the power line to be monitored to form line parameter data;

[0067] Specifically, the line electrical parameters include, but are not limited to, line resistance (line length, conductor material, cross-sectional area, etc.), line reactance (conductor arrangement), line capacitance (capacitance between the conductor and the earth), line voltage level and operation mode, etc.;

[0068] The line structure parameters include but are not limited to single-circuit lines, double-circuit lines, transformer locations, switch locations and other power facility layout parameters;

[0069] The environmental conditions of the line corridor include but are not limited to the terrain conditions along the power line (plains, mountains, etc.), vegetation coverage along the line, and distribution of buildings and facilities along the line;

[0070] Furthermore, the node load data is collected in the following manner:

[0071] Determine the important bus nodes and key user nodes in the power line; it is understandable that the important bus nodes and key user nodes are determined by professional power personnel based on common knowledge of the power grid;

[0072] Collect node load curves of important bus nodes; specifically, the node load curves include active power curves (half-hour / hour load) and reactive power curves;

[0073] Collect user electricity consumption data of key user nodes; specifically, the key user nodes may be industrial parks, large factory users or commercial buildings, etc.; the user electricity consumption data includes but is not limited to electricity consumption, electricity consumption curve, floor load distribution, load difference, etc.;

[0074] Collect the time attributes of each important load node; the time attributes include the power consumption pattern on weekdays / rest days, seasonal and holiday power consumption fluctuations; the important load nodes include important bus nodes and key user nodes;

[0075] The node load data is composed of the node load curve of the important bus nodes, the user power consumption data of the key user nodes and the time attributes of the important load nodes;

[0076] Furthermore, the device status data is collected in the following manner:

[0077] Collect the defects and fault status of each known power equipment in the power line, equipment maintenance plan and records, and basic equipment information to form equipment status data;

[0078] The defects and fault conditions include but are not limited to the degree of insulation aging, the location and type of ground faults, defects reported by online monitoring of the equipment, etc.; the equipment maintenance plan and records include regular equipment maintenance and time, temporary maintenance and emergency repair records; the basic equipment information includes but is not limited to equipment model specifications, manufacturer and service life, protection level and environmental applicable conditions, etc.;

[0079] Furthermore, the historical fault data is collected in the following manner:

[0080] Collect basic information such as the time, location, and type of power failures that occurred in the power line in the past, collect the operating status of each power equipment when each power failure occurs, collect the fault evolution process and processing records, and mark the cause of the power failure to form historical fault data;

[0081] Furthermore, the construction of the fault prediction Bayesian network model based on line auxiliary information includes the following steps:

[0082] Step 11: Based on the line parameter data and node load data, define the network nodes and their initial states of the fault prediction Bayesian network model;

[0083] Specifically, the network nodes and their initial states are defined as follows:

[0084] According to the line structure parameters in the line parameter data, create nodes representing the line, including:

[0085] For a single-circuit line, one line node is created; for a double-circuit line, two line nodes are created; the initial state of the line node is set to "normal";

[0086] According to the transformer position in the line parameter data, a transformer node representing the transformer and its initial state are created; including: creating a transformer node corresponding to the transformer device; the initial state of the transformer is obtained from the device state information of the transformer; specifically, the initial state of the transformer may be "winding normal";

[0087] According to the switch device position in the line parameter data, a switch node representing the switch and its initial switch state are created; the switch device creates a switch node correspondingly; the initial switch state is obtained from the state information of the switch device; the initial switch state can be "open" or "closed";

[0088] According to the position of each important load node in the node load data, a load node representing the load and its initial load state are created; including: creating a load node corresponding to each important load node; extracting a node load curve of the load node from the node load data, and setting the initial state of the load node according to the statistical characteristics of the node load curve;

[0089] Step 12: Based on the device status data, design the node status variables and their value ranges for each network node;

[0090] Specifically, the node state variables and their value ranges are designed as follows:

[0091] For line nodes, discrete state variables and value ranges are designed, including obtaining possible line states from equipment status information, such as "normal", "single-phase grounding", "phase-to-phase short circuit", etc., and using these possible line states as the value range of state variables of line nodes;

[0092] For transformer nodes, discrete state variables and value ranges are designed, including obtaining transformer defect types from equipment status information, such as "winding grounding" and "oil pump failure", etc.; these transformer defect types are used as the value range of the state variables of the transformer nodes;

[0093] For the "switch" node, design discrete state variables and value ranges, including: obtaining possible states of the switch from the device state information, such as "closing", "opening", "tripping", etc., and using these possible states as the discrete state variable value range of the "switch" node; further, if the switch is an intelligent switch, state values ​​such as "protection action" can be added;

[0094] Design continuous state variables and value ranges for load nodes; for example, analyze the load curve characteristics of important bus nodes from node load data, and design the value ranges of state variables such as "active power" and "reactive power" based on their statistical distribution; or analyze the power consumption curve of key user nodes and design the value range of load distribution on each floor;

[0095] Step 13: Build the topological dependency relationship between each node;

[0096] Specifically, the topological dependency relationship between the nodes is constructed in the following manner:

[0097] Using the structural information in the line parameter data, physical connection dependencies are established, for example, a line node has a dependency relationship with its connected transformer node, and a transformer node has a dependency relationship with its secondary line node;

[0098] Using the power supply relationship in the node load data, a load-power supply dependency relationship is established, for example, there is a dependency relationship between the load node and the line node to which it is connected;

[0099] Furthermore, the method of using historical fault data to train the fault prediction Bayesian network model and generate a node conditional probability table includes the following steps:

[0100] Step 21: Obtain the node status of each node at each power failure from the historical fault case data to obtain the fault state vector at each power failure;

[0101] An example of a simple fault state vector is:

[0102] Assume that our network has three nodes A, B, and C, and they have the following states: node A: faulty, node B: overloaded, node C: grounded; then a historical case "node A faulty, node B normal, node C grounded" can be represented as [1,0,2], where 1 indicates that node A is faulty, 0 indicates that node B is normal, and 2 indicates that node C is grounded;

[0103] Step 22: Input each fault state vector into the fault prediction Bayesian network model according to the corresponding relationship of the nodes, perform CPT parameter estimation on the CPT parameters of each node, and train the model using parameter learning and structure optimization algorithms until the CPT parameter estimation value of the node reaches convergence and the training is stopped when the optimal network model structure is generated;

[0104] Step 23: Read the CPT parameters of each node to obtain the final node conditional probability table; it should be noted that, for each node, the CPT parameters are the frequency / conditional probability of each state combination.

[0105] A simple example of a CPT parameter is as follows:

[0106] There is a power system Bayesian network model with 2 nodes, including line node (L) and transformer node (T). Their state variables and value ranges are as follows:

[0107] Line node L: state = {normal, single-phase grounding, phase-to-phase short circuit};

[0108] Transformer node T: state = {normal, winding grounded};

[0109] The CPT parameters of the line node L can be:

[0110] Normal 0.950.040.01

[0111] Winding grounding 0.750.200.05

[0112] The first row indicates the probability of line L being normal / single-phase grounding / phase-to-phase short circuit when transformer T is normal;

[0113] The second row indicates the probability of line L being normal / single-phase grounded / phase-to-phase short circuit when the transformer T winding is grounded;

[0114] Furthermore, the method of collecting real-time monitoring data of the power line is:

[0115] Collect real-time node load data and real-time equipment status information of power lines during real-time operation to form real-time monitoring data;

[0116] It can be understood that the real-time node load data and real-time device status information are collected in real time by collecting the real-time node state variables of each load node to form the real-time node load data, and the node state variables of each power equipment node (transformer, switch, line, etc.) to form the real-time device status information;

[0117] Furthermore, the method of mapping the real-time monitoring data into network node evidence is:

[0118] Converting the node state variables of each network node in the real-time monitoring data into a fault location feature vector consistent with the form of the fault state vector, wherein the fault location feature vector is the network node evidence;

[0119] Furthermore, the method of inputting the network node evidence obtained in real time into the fault prediction Bayesian network model and calculating the fault probability of each network node based on the node conditional probability table is as follows:

[0120] Input the fault location feature vector into the fault prediction Bayesian network model;

[0121] Determine the Markov spiral solution order of each network node according to the network topology of the fault prediction Bayesian network model;

[0122] Traverse each network node in order, query the conditional probability from the node conditional probability table of its parent node, and use it as the fault probability of the network node; it can be understood that the fault probability is obtained by reasoning based on the status of other network nodes in the network topology structure, thereby reducing the risk of errors in the status provided by the network node and improving the accuracy of power fault location;

[0123] A simple example of calculating the probability of failure is as follows:

[0124] When the input fault location feature vector is [1,0], (1 indicates that the line node L is single-phase grounded, and 0 indicates that the transformer node T is winding grounded;)

[0125] The process of probabilistic reasoning using CPT is:

[0126] Query from the node conditional probability table of the transformer node:

[0127] P(T=normal|L=single-phase grounding)=0.75;

[0128] P(T=winding grounding|L=single-phase grounding)=0.25;

[0129] Thus P(T=Normal)=0.75

[0130] P(T=winding grounded)=0.25;

[0131] Furthermore, the method of selecting nodes with high abnormal probability as fault troubleshooting nodes is as follows:

[0132] A fault probability threshold is set in advance, and all network nodes whose fault probabilities are greater than the fault probability threshold are selected as fault troubleshooting nodes, and the fault troubleshooting nodes are sorted from large to small according to the fault probability.

[0133] Example 2

[0134] like Figure 2 As shown, the intelligent fault location system based on multi-source power information integration includes a basic data collection module, a Bayesian network model training module and an intelligent fault location module; wherein each module is electrically connected;

[0135] The basic data collection module collects line parameter data, node load data and equipment status information in advance to form line auxiliary information, and collects historical fault data in advance and sends the historical fault data to the Bayesian network model training module;

[0136] The Bayesian network model training module builds a fault prediction Bayesian network model based on line auxiliary information, and uses historical fault data to train the fault prediction Bayesian network model, generates a node conditional probability table, and sends the fault prediction Bayesian network model and the node conditional probability table to the intelligent fault location module;

[0137] The intelligent fault location module collects real-time monitoring data of power lines and maps the real-time monitoring data into network node evidence. It inputs the real-time acquired network node evidence into the fault prediction Bayesian network model, calculates the fault probability of each network node based on the node conditional probability table, and selects nodes with high abnormal probability as fault detection nodes.

[0138] Example 3

[0139] Figure 3 1 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 3 As shown, according to another aspect of the present application, an electronic device 100 is also provided. The electronic device 100 may include one or more processors and one or more memories. The memories store computer readable codes, and when the computer readable codes are executed by one or more processors, the intelligent fault location method based on multi-source power information integration as described above may be executed.

[0140] The method or device according to the embodiment of the present application can also be used by Figure 3 The electronic device architecture shown in FIG. Figure 3 As shown, the electronic device 100 may include a bus 101, one or more CPUs 102, a ROM 103, a RAM 104, a communication port 105 connected to a network, an input / output component 106, a hard disk 107, etc. The storage device in the electronic device 100, such as the ROM 103 or the hard disk 107, may store the intelligent fault location method based on multi-source power information integration provided in the present application.

[0141] Furthermore, the electronic device 100 may further include a user interface 108. Figure 3 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 3 One or more components of an electronic device are shown.

[0142] Example 4

[0143] Figure 4 Schematic diagram of a computer-readable storage medium structure provided by an embodiment of the present application. Figure 4As shown, a computer-readable storage medium 200 according to an embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 200. When the computer-readable instructions are executed by the processor, the intelligent fault location method based on multi-source power information integration according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium 200 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0144] In addition, according to the implementation of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, the non-transitory machine-readable storage medium stores machine-readable instructions, the machine-readable instructions can be run by a processor to execute instructions corresponding to the method steps provided by the present application, and when the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0145] The methods, apparatuses, and devices of the present application may be implemented in many ways. For example, the methods, apparatuses, and devices of the present application may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present application are not limited to the order specifically described above unless otherwise specified. In addition, in some embodiments, the present application may also be implemented as programs recorded in a recording medium, which include machine-readable instructions for implementing the method according to the present application. Therefore, the present application also covers recording media storing programs for executing the method according to the present application.

[0146] In addition, the parts of the above-mentioned technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0147] The specific implementation modes as described above further describe the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation mode of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0148] The above preset parameters or preset thresholds are all set by technicians in this field according to actual conditions or obtained through large-scale data simulation.

[0149] 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. An intelligent fault location method based on multi-source power information integration, characterized in that: The following steps are involved: Step 1: Collect line parameter data, node load data and equipment status information in advance to form line auxiliary information, and collect historical fault data in advance; Step 2: Build a fault prediction Bayesian network model based on line auxiliary information, and use historical fault data to train the fault prediction Bayesian network model to generate a node conditional probability table; Step 3: Collect real-time monitoring data of power lines and map the real-time monitoring data into network node evidence; Step 4: Input the real-time acquired network node evidence into the fault prediction Bayesian network model, calculate the fault probability of each network node based on the node conditional probability table, and select nodes with high abnormal probability as fault troubleshooting nodes; The node load data is collected in the following manner: Determine important bus nodes and key user nodes in power lines; Collect node load curves of important busbar nodes; Collect user electricity consumption data of key user nodes; Collecting the time attributes of each important load node; the important load nodes include important bus nodes and key user nodes; The node load data is composed of the node load curve of the important bus nodes, the user power consumption data of the key user nodes and the time attributes of the important load nodes; The device status information is collected in the following manner: Collect the defects and fault status of each known power equipment in the power line, equipment maintenance plan and records, and basic equipment information to form equipment status data; The historical fault data is collected in the following manner: Collect basic information about past power failures in power lines, collect the operating status of each power device when each power failure occurs, collect the failure evolution process and processing records, and mark the cause of the power failure to form historical failure data.

2. The intelligent fault location method based on multi-source power information integration according to claim 1 is characterized in that: The line parameter data is collected in the following manner: The line electrical parameters, line structural parameters and line corridor environmental conditions of the power line to be monitored are collected to form line parameter data.

3. The intelligent fault location method based on multi-source power information integration according to claim 2 is characterized in that: The construction of the fault prediction Bayesian network model based on line auxiliary information includes the following steps: Step 11: Based on the line parameter data and node load data, define the network nodes and their initial states of the fault prediction Bayesian network model; Step 12: Based on the device status data, design the node status variables and their value ranges for each network node; Step 13: Build the topological dependency relationship between each node; The topological dependency relationship between the nodes is constructed in the following manner: Using the structural information in the line parameter data, physical connection dependencies are established; The load-power supply dependency is established by utilizing the power supply relationship in the node load data.

4. The intelligent fault location method based on multi-source power information integration according to claim 3 is characterized in that: The method of using historical fault data to train the fault prediction Bayesian network model and generate a node conditional probability table comprises the following steps: Step 21: Obtain the node status of each node at each power failure from the historical fault case data to obtain the fault state vector at each power failure; Step 22: Input each fault state vector into the fault prediction Bayesian network model according to the corresponding relationship of the nodes, perform CPT parameter estimation on the CPT parameters of each node, and train the model using parameter learning and structure optimization algorithms until the CPT parameter estimation value of the node reaches convergence and the training is stopped when the optimal network model structure is generated; Step 23: Read the CPT parameters of each node to obtain the final node conditional probability table.

5. The intelligent fault location method based on multi-source power information integration according to claim 4 is characterized in that: The method of inputting the network node evidence acquired in real time into the fault prediction Bayesian network model and calculating the fault probability of each network node based on the node conditional probability table is as follows: Input the fault location feature vector into the fault prediction Bayesian network model; Determine the Markov spiral solution order of each network node according to the network topology of the fault prediction Bayesian network model; Traverse each network node in order, query the conditional probability from the node conditional probability table of its parent node as the failure probability of the network node.

6. The intelligent fault location method based on multi-source power information integration according to claim 5 is characterized in that: The method of selecting nodes with high abnormal probability as fault troubleshooting nodes is as follows: Set the fault probability threshold in advance and set all network nodes whose fault probability is greater than the fault probability threshold. Points are selected as troubleshooting nodes, and the troubleshooting nodes are sorted from large to small according to the failure probability.

7. An intelligent fault location system based on multi-source power information integration, which is used to implement the intelligent fault location method based on multi-source power information integration as described in any one of claims 1 to 6, characterized in that: It includes a basic data collection module, a Bayesian network model training module and an intelligent fault location module; wherein each module is electrically connected; The basic data collection module collects line parameter data, node load data and equipment status information in advance to form line auxiliary information, and collects historical fault data in advance and sends the historical fault data to the Bayesian network model training module; The Bayesian network model training module builds a fault prediction Bayesian network model based on line auxiliary information, and uses historical fault data to train the fault prediction Bayesian network model, generates a node conditional probability table, and sends the fault prediction Bayesian network model and the node conditional probability table to the intelligent fault location module; The intelligent fault location module collects real-time monitoring data of power lines and maps the real-time monitoring data into network node evidence. It inputs the real-time acquired network node evidence into the fault prediction Bayesian network model, calculates the fault probability of each network node based on the node conditional probability table, and selects nodes with high abnormal probability as fault detection nodes.

8. An electronic device, characterized in that: include: processor and memory, wherein The memory stores a computer program that can be called by the processor; The processor executes the intelligent fault location method based on multi-source power information integration as described in any one of claims 1-6 in the background by calling the computer program stored in the memory.

9. A computer-readable storage medium, characterized in that: A rewritable computer program is stored thereon; When the computer program is executed on a computer device, the computer device executes the intelligent fault location method based on multi-source power information integration as described in any one of claims 1 to 6 in the background.

Images