Method and system for positioning and diagnosing fault section of power distribution network based on artificial intelligence

Abstract

The invention relates to the technical field of power distribution networks, in particular to a power distribution network fault section positioning diagnosis method and system based on artificial intelligence. Comprising the steps of setting a plurality of monitoring points based on equipment parameters of a power distribution network, and generating a plurality of power distribution sections according to all the monitoring points; a diagnosis agent is constructed according to all the power distribution sections, and the diagnosis agent generates a fault positioning result based on the obtained fault feedback packet; obtaining a diagnosis record packet according to the preset update time node, and judging whether an update instruction is generated or not according to the diagnosis record packet; setting a plurality of monitoring points based on structural parameters of the power distribution network, performing segmentation processing on the power distribution network through all the monitoring points to generate a plurality of power distribution sections, analyzing the operation state of each monitoring point through F, quickly screening abnormal sections which may have fault risks, and performing fault diagnosis on the analysis result of each abnormal section through a fault diagnosis model. Fault sections are quickly positioned, and the inspection efficiency of the power distribution network is improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of power distribution networks, in particular to a power distribution network fault section positioning diagnosis method and system based on artificial intelligence. BACKGROUND

[0002] As the final link of the power system facing users, the power supply reliability of the power distribution network is crucial. Rapid and accurate positioning of the fault section is the key to narrowing the power outage range and improving the power restoration efficiency.

[0003] The existing power distribution network fault positioning method relies heavily on preset and fixed power grid topology and reasoning rules. When the power distribution network structure frequently changes due to operation mode adjustment or distributed energy access, its adaptability is poor, and the fault tolerance is low, and errors are easily caused by abnormal communication of terminal equipment (FTU / DTU) or information distortion. SUMMARY

[0004] The application aims to solve the above technical problems. The application provides a power distribution network fault section positioning diagnosis method and system based on artificial intelligence, aiming to improve the fault positioning accuracy and inspection efficiency of the power distribution network.

[0005] In some embodiments of the application, a plurality of monitoring points are set based on the structural parameters of the power distribution network, and the power distribution network is segmented by all the monitoring points to generate a plurality of power distribution sections. The operating state of each monitoring point is analyzed by F to quickly screen abnormal sections that may have a fault risk, and the analysis result of each abnormal section is analyzed by a fault diagnosis model to quickly locate the fault section, thereby improving the inspection efficiency of the power distribution network.

[0006] In some embodiments of the application, a double diagnosis model is constructed. On the one hand, the operating state of each monitoring point is analyzed to achieve primary screening of abnormal areas, and on the other hand, a fault positioning model is constructed to realize double verification of the fault section and the abnormal state of the monitoring point, thereby improving the fault positioning accuracy of the power distribution network.

[0007] In some embodiments of the application, a power distribution network fault section positioning diagnosis method based on artificial intelligence is provided, which comprises:

[0008] A plurality of monitoring points are set based on the device parameters of the power distribution network, and a plurality of power distribution sections are generated according to all the monitoring points;

[0009] A diagnosis intelligent agent is constructed according to all the power distribution sections, and the diagnosis intelligent agent generates a fault positioning result based on the obtained fault feedback package;

[0010] A diagnosis record package is obtained according to a preset update time node, and it is judged whether an update instruction is generated according to the diagnosis record package;

[0011] Wherein, a plurality of distribution section is set, including:

[0012] Establish a distribution section sequence A, A=(a1, a2…an), wherein ai is the ith distribution section; n is the number of distribution sections. i …an n ), wherein ai is the ith distribution section; n is the number of distribution sections. i

[0013] In some embodiments of the application, a diagnostic agent is constructed, including:

[0014] Establish a monitoring point sequence B, B=(b1, b2…bm), wherein bi is the ith monitoring point; m is the number of monitoring points; i …bm m ), wherein bi is the ith monitoring point; m is the number of monitoring points. i

[0015] According to the monitoring point sequence B, bi is set as the target monitoring point in turn;

[0016] Generate the correlation evaluation value of the target monitoring point and each distribution section;

[0017] According to all correlation evaluation values, set the section mapping table of the target monitoring point;

[0018] Construct the evaluation sub-model of the target monitoring point;

[0019] According to the section mapping table and the evaluation sub-model, generate the diagnostic sub-model of the target monitoring point;

[0020] In turn, generate the diagnostic sub-model of each monitoring point;

[0021] According to all diagnostic sub-models, generate the fault diagnosis model.

[0022] In some embodiments of the application, the diagnostic agent is constructed, further including:

[0023] According to the distribution section sequence A, ai is set as the target section in turn; i

[0024] Obtain the historical fault package of the target section;

[0025] Generate the mapping evaluation value between the target section and each monitoring point;

[0026] Establish a mapping evaluation value sequence C, C=(c1, c2…cm), wherein ci is the mapping evaluation value between the target section and the ith monitoring point; m is the number of monitoring points; i …cm m ), wherein ci is the mapping evaluation value between the target section and the ith monitoring point; m is the number of monitoring points. i

[0027] Pre-set the mapping evaluation value threshold C1;

[0028] If ci is less than C1, the target section is determined as the target section of the ith monitoring point. i ​​​​>C1, set the i-th monitoring point as the mapping monitoring point of the target segment;

[0029] Obtain all mapped monitoring points in the target area;

[0030] Set the mapping sub-state of each mapping monitoring point based on historical fault packets;

[0031] Construct a localization sub-model for the target segment based on all mapping sub-states;

[0032] The location sub-models for each power distribution section are set sequentially;

[0033] Generate a fault location model based on all location sub-models;

[0034] A diagnostic intelligent agent is constructed based on the fault diagnosis model and the fault location model.

[0035] In some embodiments of this application, generating a mapping evaluation value between the target segment and each monitoring point includes:

[0036] Based on the monitoring point sequence B, bi is sequentially set as the monitoring point to be compared;

[0037] Generate a mapping evaluation value c between the monitoring point to be compared and the target section;

[0038]

[0039] Where θ1 is the number of fault records in the historical fault packets of the target segment; s i The first-level fluctuation value of the monitoring point to be compared in the i-th fault record;

[0040] The mapping evaluation values ​​between the target section and each monitoring point are generated sequentially.

[0041] In some embodiments of this application, the generation of fault location results includes:

[0042] b is set sequentially according to the monitoring point sequence B. i These are monitoring points to be evaluated.

[0043] Generate sub-feedback packets for the monitoring points to be evaluated based on the fault feedback packets;

[0044] The diagnostic sub-model for the monitoring points to be evaluated is set as a primary diagnostic model;

[0045] Generate the abnormal risk value f of the monitoring point to be evaluated based on the primary diagnostic model and sub-feedback package;

[0046] Preset anomaly risk threshold F1;

[0047] If f > F1, obtain all associated segments of the monitoring point to be evaluated according to the primary diagnostic model;

[0048] Each associated segment of the monitoring point to be evaluated is designated as an abnormal segment;

[0049] The abnormal risk values ​​for each monitoring point are generated sequentially.

[0050] Establish an abnormal segment sequence A1 based on all abnormal risk values, where A1 = (a 11 a 12 …a 1i …a 1n1 ), where a 1i Let n1 be the i-th abnormal segment; n1 is the number of abnormal segments.

[0051] Fault location results are generated based on all abnormal sections.

[0052] In some embodiments of this application, generating the abnormal risk value f of the monitoring point to be evaluated includes:

[0053]

[0054] Where θ2 is the number of monitoring indicators for the monitoring points to be evaluated; β i v is the influence factor of the i-th monitoring indicator among the monitoring points to be evaluated; i It is the first-level reference value of the i-th monitoring indicator among the monitoring points to be evaluated, generated based on the sub-feedback package; v' i It is the standard reference value of the i-th monitoring indicator among the monitoring points to be evaluated, generated based on the primary diagnostic model.

[0055] In some embodiments of this application, fault location results are generated based on all abnormal segments, including:

[0056] Based on the abnormal segment sequence A1, a is set sequentially. 1i The target abnormal section;

[0057] The target anomaly segment is defined as the target localization model;

[0058] The matching evaluation value d of the target abnormal section is generated based on the target localization model and the fault feedback package;

[0059] Consistently generate the matching evaluation values ​​for each abnormal segment;

[0060] Establish a sequence of matching evaluation values ​​D, where D = (d1, d2, ..., dn). i …d n1 ), where d i is the matching evaluation value of the i-th abnormal segment; n1 is the number of abnormal segments;

[0061] The preset matching evaluation threshold D1;

[0062] Obtain the maximum value d in the sequence of matching evaluation values ​​D.max ;

[0063] If d max >D1, generate first-level positioning instructions

[0064] If d max <D1, set the maximum value d max The corresponding abnormal section is the target fault section;

[0065] The first-level positioning instructions include:

[0066] If di > d max， The i-th abnormal segment is set as the target fault segment.

[0067] In some embodiments of this application, generating the matching evaluation value d of the target abnormal segment includes:

[0068]

[0069] Where θ3 is the number of mapping monitoring points in the target localization model; η i k represents the influence factor of the i-th mapping monitoring point in the target localization model. i To generate a similarity value for the i-th mapped monitoring point based on the fault feedback packet.

[0070] In some embodiments of this application, an artificial intelligence-based distribution network fault location and diagnosis system is provided, comprising:

[0071] The central control unit is used to set multiple monitoring points according to the equipment parameters of the power distribution network;

[0072] The central control unit is also used to generate multiple power distribution sections based on all monitoring points;

[0073] The monitoring unit includes multiple monitoring sub-modules, which are located at various monitoring points.

[0074] The monitoring unit is used to collect operational data from each monitoring point;

[0075] The monitoring unit is also used to generate fault feedback packages;

[0076] The central control unit includes:

[0077] The first processing module is used to construct a diagnostic intelligent agent based on all power distribution sections. The diagnostic intelligent agent generates fault location results based on the acquired fault feedback packets.

[0078] The second processing module is used to obtain the diagnostic record package according to the preset update time node, and determine whether to generate an update instruction based on the diagnostic record package.

[0079] The third processing module is used to establish the power distribution section sequence A, A = (a1, a2, ..., a... i …a n ), where a i Let be the i-th power distribution section; n is the number of power distribution sections.

[0080] In some embodiments of this application, the first processing module is further configured to:

[0081] Establish a sequence of monitoring points B, B = (b1, b2, ..., bb2) i …b m ), where b i Let m be the i-th monitoring point; m is the number of monitoring points.

[0082] Based on the monitoring point sequence B, bi is sequentially set as the target monitoring point;

[0083] Generate correlation evaluation values ​​between target monitoring points and each power distribution section;

[0084] A segment mapping table for target monitoring points is set based on all associated evaluation values;

[0085] Construct an evaluation sub-model for the target monitoring points;

[0086] A diagnostic sub-model for the target monitoring point is generated based on the segment mapping table and the evaluation sub-model.

[0087] Diagnostic sub-models for each monitoring point are generated sequentially.

[0088] Generate a fault diagnosis model based on all diagnostic sub-models;

[0089] Based on the power distribution section sequence A, a is set sequentially. i For the target section;

[0090] Obtain historical fault packets for the target segment;

[0091] Generate a mapping evaluation value between the target section and each monitoring point;

[0092] Establish a mapping evaluation value sequence C, C = (c1, c2, ..., c3) i …c m ), where c i is the mapping evaluation value between the target segment and the i-th monitoring point; m is the number of monitoring points;

[0093] Preset mapping evaluation value threshold C1;

[0094] If c i >C1, set the i-th monitoring point as the mapping monitoring point of the target segment;

[0095] Obtain all mapped monitoring points in the target area;

[0096] Set the mapping sub-state of each mapping monitoring point based on historical fault packets;

[0097] Construct a localization sub-model for the target segment based on all mapping sub-states;

[0098] The location sub-models for each power distribution section are set sequentially;

[0099] Generate a fault location model based on all location sub-models;

[0100] A diagnostic intelligent agent is constructed based on the fault diagnosis model and the fault location model.

[0101] Compared with existing technologies, the advantages of the artificial intelligence-based method and system for locating and diagnosing fault sections in power distribution networks disclosed in this application are as follows:

[0102] Multiple monitoring points are set based on the structural parameters of the distribution network, and the distribution network is divided into multiple distribution sections through all monitoring points. The operating status of each monitoring point is analyzed by F to quickly screen abnormal sections that may have fault risks. The fault diagnosis model is used to analyze the results of each abnormal section to quickly locate the fault section and improve the inspection efficiency of the distribution network.

[0103] By constructing a dual diagnostic model, on the one hand, the operation status of each monitoring point is analyzed to achieve initial screening of abnormal areas; on the other hand, a fault location model is constructed to achieve dual verification of the fault section and the abnormal status of the monitoring point, thereby improving the fault location accuracy of the distribution network. Attached Figure Description

[0104] Figure 1 This is a flowchart illustrating a preferred embodiment of an artificial intelligence-based method for locating and diagnosing fault sections in a power distribution network. Detailed Implementation

[0105] The specific embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate this application, but are not intended to limit the scope of this application.

[0106] In the description of this application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0107] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0108] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0109] like Figure 1 As shown in the preferred embodiment of this application, a method for locating and diagnosing fault sections in a distribution network based on artificial intelligence includes:

[0110] S101: Based on the equipment parameters of the power distribution network, set multiple monitoring points and generate multiple power distribution sections based on all monitoring points;

[0111] S102: Construct a diagnostic intelligent agent based on all power distribution sections. The diagnostic intelligent agent generates fault location results based on the acquired fault feedback packets.

[0112] S103: Obtain the diagnostic record package according to the preset update time node, and determine whether to generate an update instruction based on the diagnostic record package;

[0113] This includes setting up multiple order allocation sections, including:

[0114] Establish a power distribution section sequence A, A = (a1, a2, ..., a3) i …a n ), where a i Let be the i-th power distribution section; n is the number of power distribution sections.

[0115] Specifically, the structural parameters of the distribution network are traversed to set each equipment point (such as circuit breakers, sectionalizing switches, distribution transformers, and other non-cable structures) as monitoring points.

[0116] Specifically, the power distribution network is divided according to all monitoring points. The cable structure between any two monitoring points is a power distribution section, and all power distribution sections are generated in sequence.

[0117] Specifically, there are two monitoring points at both ends of a single power distribution section.

[0118] It is understood that in the above embodiments, multiple monitoring points are set based on the structural parameters of the distribution network, and the distribution network is divided into multiple distribution sections through all monitoring points. The operating status of each monitoring point is analyzed by F to quickly screen abnormal sections that may have fault risks. The fault diagnosis model is used to analyze the results of each abnormal section to quickly locate the fault section and improve the inspection efficiency of the distribution network.

[0119] In a preferred embodiment of this application, the construction of a diagnostic intelligent agent includes:

[0120] Establish a sequence of monitoring points B, B = (b1, b2, ..., bb2) i …b m ), where b i Let m be the i-th monitoring point; m is the number of monitoring points.

[0121] Based on the monitoring point sequence B, bi is sequentially set as the target monitoring point;

[0122] Generate correlation evaluation values ​​between target monitoring points and each power distribution section;

[0123] A segment mapping table for target monitoring points is set based on all associated evaluation values;

[0124] Construct an evaluation sub-model for the target monitoring points;

[0125] A diagnostic sub-model for the target monitoring point is generated based on the segment mapping table and the evaluation sub-model.

[0126] Diagnostic sub-models for each monitoring point are generated sequentially.

[0127] A fault diagnosis model is generated based on all the diagnostic sub-models.

[0128] Specifically, if the target monitoring point is the endpoint of the current power distribution section, the correlation evaluation value between the two is 1, and the current power distribution section is set as the associated section of the target monitoring point; if the target monitoring point is not the endpoint of the current power distribution section, the correlation evaluation value between the two is 0, and the current power distribution section is set as the non-associated section of the target monitoring point.

[0129] Specifically, the target monitoring point's segment mapping table includes all associated segments of the target monitoring point.

[0130] Specifically, based on the equipment category of the target monitoring point, corresponding monitoring indicators are set. These indicators include, but are not limited to, parameters that can characterize the operating status of the equipment corresponding to the monitoring point, such as the effective value / transient waveform of three-phase current and voltage.

[0131] Specifically, after quantifying all monitoring indicators of the target monitoring point, each indicator is within the same value range. Standard reference values ​​are set for each monitoring indicator of the target monitoring point under normal conditions. An evaluation sub-model of the target monitoring point is constructed based on all standard reference values. The evaluation sub-model represents the normal operating state of the target monitoring point when no faults occur in all power distribution sections.

[0132] In a preferred embodiment of this application, constructing a diagnostic intelligent agent further includes:

[0133] Based on the power distribution section sequence A, a is set sequentially. i For the target section;

[0134] Obtain historical fault packets for the target segment;

[0135] Generate a mapping evaluation value between the target section and each monitoring point;

[0136] Establish a mapping evaluation value sequence C, C = (c1, c2, ..., c3) i …c m ), where c i is the mapping evaluation value between the target segment and the i-th monitoring point; m is the number of monitoring points;

[0137] Preset mapping evaluation value threshold C1;

[0138] If c i >C1, set the i-th monitoring point as the mapping monitoring point of the target segment;

[0139] Obtain all mapped monitoring points in the target area;

[0140] Set the mapping sub-state of each mapping monitoring point based on historical fault packets;

[0141] Construct a localization sub-model for the target segment based on all mapping sub-states;

[0142] The location sub-models for each power distribution section are set sequentially;

[0143] Generate a fault location model based on all location sub-models;

[0144] A diagnostic intelligent agent is constructed based on the fault diagnosis model and the fault location model.

[0145] Specifically, the data in the historical fault package includes all fault records of the target section and the historical status of each monitoring point at each time a fault occurs.

[0146] Specifically, the mapping evaluation value can be set based on historical parameters. When the real-time mapping evaluation value is greater than the mapping evaluation value threshold, it indicates that when a fault occurs in the target section, the current monitoring point will experience state fluctuations.

[0147] Specifically, by analyzing the state fluctuation parameters of the current mapping monitoring point when a fault occurs in the target section, a corresponding mapping sub-state is constructed. The mapping sub-state represents the real-time operating status of the current mapping monitoring point when a fault occurs in the target section.

[0148] Specifically, generating mapping evaluation values ​​between the target section and each monitoring point includes:

[0149] Based on the monitoring point sequence B, bi is sequentially set as the monitoring point to be compared;

[0150] Generate a mapping evaluation value c between the monitoring point to be compared and the target section;

[0151]

[0152] Where θ1 is the number of fault records in the historical fault packets of the target segment; s i The first-level fluctuation value of the monitoring point to be compared in the i-th fault record;

[0153] The mapping evaluation values ​​between the target section and each monitoring point are generated sequentially.

[0154] Specifically, the state constructed in the evaluation sub-model of the monitoring point to be compared is compared with the state when the fault occurs in the target section. The corresponding first-level fluctuation value is set according to the degree of difference. The greater the degree of difference, the greater the corresponding first-level fluctuation value. The mapping relationship between the two can be set according to historical parameters.

[0155] Specifically, the larger the mapping evaluation value, the greater the impact on the monitoring point to be compared when a fault occurs in the target section, meaning that the operating status of the monitoring point to be compared is more likely to fluctuate.

[0156] In a preferred embodiment of this application, generating fault location results includes:

[0157] b is set sequentially according to the monitoring point sequence B. i These are monitoring points to be evaluated.

[0158] Generate sub-feedback packets for the monitoring points to be evaluated based on the fault feedback packets;

[0159] The diagnostic sub-model for the monitoring points to be evaluated is set as a primary diagnostic model;

[0160] Generate the abnormal risk value f of the monitoring point to be evaluated based on the primary diagnostic model and sub-feedback package;

[0161] Preset anomaly risk threshold F1;

[0162] If f > F1, obtain all associated segments of the monitoring point to be evaluated according to the primary diagnostic model;

[0163] Each associated segment of the monitoring point to be evaluated is designated as an abnormal segment;

[0164] The abnormal risk values ​​for each monitoring point are generated sequentially.

[0165] Establish an abnormal segment sequence A1 based on all abnormal risk values, where A1 = (a 11 a 12 …a 1i …a 1n1 ), where a 1i Let n1 be the i-th abnormal segment; n1 is the number of abnormal segments.

[0166] Fault location results are generated based on all abnormal sections.

[0167] Specifically, the abnormal risk value threshold can be set based on historical parameters. If the abnormal risk value of the monitoring point to be evaluated is greater than the preset abnormal risk value threshold, it indicates that there is abnormal fluctuation when the monitoring point to be evaluated experiences a fault, that is, the fault occurs within the associated section of the monitoring point to be evaluated.

[0168] Specifically, by judging whether there are abnormal states at each monitoring point, relevant power distribution sections that may have fault risks can be quickly screened, thereby establishing abnormal power distribution sections.

[0169] Specifically, generating the abnormal risk value f for the monitoring points to be evaluated includes:

[0170]

[0171] Where θ2 is the number of monitoring indicators for the monitoring points to be evaluated; β i v is the influence factor of the i-th monitoring indicator among the monitoring points to be evaluated; i It is the first-level reference value of the i-th monitoring indicator among the monitoring points to be evaluated, generated based on the sub-feedback package; v' i It is the standard reference value of the i-th monitoring indicator among the monitoring points to be evaluated, generated based on the primary diagnostic model.

[0172] Specifically, the influence factors of each monitoring indicator are set according to the degree of mapping of the monitoring point to be evaluated. The greater the degree of mapping, the larger the corresponding shadow factor.

[0173] Specifically, the first-level reference value is the real-time value of each monitoring indicator generated based on the relevant data in the sub-feedback package.

[0174] Specifically, the higher the abnormal risk value, the greater the fluctuation in the operating status of the monitoring point to be evaluated when a fault occurs, and the greater the possibility of fault risk in each related section of the monitoring point to be evaluated.

[0175] It is understandable that in the above embodiments, by analyzing the fault feedback packets, it is determined whether there are abnormal fluctuations at each monitoring point. Based on the analysis results, multiple abnormal segments are screened to reduce the scope of subsequent fault location and improve fault location efficiency.

[0176] In some embodiments of this application, fault location results are generated based on all abnormal segments, including:

[0177] Based on the abnormal segment sequence A1, a is set sequentially. 1i The target abnormal section;

[0178] The target anomaly segment is defined as the target localization model;

[0179] The matching evaluation value d of the target abnormal section is generated based on the target localization model and the fault feedback package;

[0180] Consistently generate the matching evaluation values ​​for each abnormal segment;

[0181] Establish a sequence of matching evaluation values ​​D, where D = (d1, d2, ..., dn). i …d n1 ), where d i is the matching evaluation value of the i-th abnormal segment; n1 is the number of abnormal segments;

[0182] The preset matching evaluation threshold D1;

[0183] Obtain the maximum value d in the sequence of matching evaluation values ​​D. max ;

[0184] If d max >D1, generate first-level positioning instructions

[0185] If d max <D1, set the maximum value d max The corresponding abnormal section is the target fault section;

[0186] The first-level positioning instructions include:

[0187] If di > d max， The i-th abnormal segment is set as the target fault segment.

[0188] Specifically, the matching evaluation value threshold can be set based on historical parameters. When the matching evaluation value is greater than the preset matching evaluation value threshold, it indicates that the data in the fault feedback package matches the various characteristics of the target abnormal section when a fault occurs, and there is an operational fault in the target abnormal area.

[0189] Specifically, the matching evaluation value d for the target anomaly segment includes:

[0190]

[0191] Where θ3 is the number of mapping monitoring points in the target localization model; η i k represents the influence factor of the i-th mapping monitoring point in the target localization model. i To generate a similarity value for the i-th mapped monitoring point based on the fault feedback packet.

[0192] Specifically, the higher the fit evaluation value, the greater the possibility that there is a fault in the target abnormal section.

[0193] Specifically, the real-time operating status of the current mapping monitoring point is constructed based on the fault feedback package, and a corresponding similarity value is set according to the difference between the real-time operating status and the mapping sub-state of the current mapping monitoring point in the target positioning model. The smaller the difference, the larger the corresponding similarity value. The mapping relationship between the two can be set according to historical parameters.

[0194] Specifically, the influence factor of each mapping monitoring point is set according to its mapping evaluation value with the target abnormal section. The larger the mapping evaluation value, the larger the corresponding influence factor.

[0195] Specifically, the location results are generated based on all selected target fault sections, and maintenance personnel formulate corresponding maintenance plans based on the location results.

[0196] In another preferred embodiment of the artificial intelligence-based distribution network fault section location and diagnosis method based on any of the above preferred embodiments, this preferred embodiment provides an artificial intelligence-based distribution network fault section location and diagnosis system, including:

[0197] The central control unit is used to set multiple monitoring points according to the equipment parameters of the power distribution network;

[0198] The central control unit is also used to generate multiple power distribution sections based on all monitoring points;

[0199] The monitoring unit includes multiple monitoring sub-modules, which are set up at various monitoring points;

[0200] The monitoring unit is used to collect operational data from each monitoring point;

[0201] The monitoring unit is also used to generate fault feedback packets;

[0202] The central control unit includes:

[0203] The first processing module is used to construct a diagnostic intelligent agent based on all power distribution sections. The diagnostic intelligent agent generates fault location results based on the acquired fault feedback packets.

[0204] The second processing module is used to obtain the diagnostic record package according to the preset update time node, and determine whether to generate an update instruction based on the diagnostic record package.

[0205] The third processing module is used to establish the power distribution section sequence A, A = (a1, a2, ..., a... i …a n ), where a i Let be the i-th power distribution section; n is the number of power distribution sections.

[0206] Specifically, the monitoring submodule is preferably a data acquisition device of various types.

[0207] In some embodiments of this application, the first processing module is further configured to:

[0208] Establish a sequence of monitoring points B, B = (b1, b2, ..., bb2) i …b m ), where b i Let m be the i-th monitoring point; m is the number of monitoring points.

[0209] Based on the monitoring point sequence B, bi is sequentially set as the target monitoring point;

[0210] Generate correlation evaluation values ​​between target monitoring points and each power distribution section;

[0211] A segment mapping table for target monitoring points is set based on all associated evaluation values;

[0212] Construct an evaluation sub-model for the target monitoring points;

[0213] A diagnostic sub-model for the target monitoring point is generated based on the segment mapping table and the evaluation sub-model.

[0214] Diagnostic sub-models for each monitoring point are generated sequentially.

[0215] Generate a fault diagnosis model based on all diagnostic sub-models;

[0216] Based on the power distribution section sequence A, a is set sequentially. i For the target section;

[0217] Obtain historical fault packets for the target segment;

[0218] Generate a mapping evaluation value between the target section and each monitoring point;

[0219] Establish a mapping evaluation value sequence C, C = (c1, c2, ..., c3) i …c m ), where ci is the mapping evaluation value between the target segment and the i-th monitoring point; m is the number of monitoring points;

[0220] Preset mapping evaluation value threshold C1;

[0221] If c i >C1, set the i-th monitoring point as the mapping monitoring point of the target segment;

[0222] Obtain all mapped monitoring points in the target area;

[0223] Set the mapping sub-state of each mapping monitoring point based on historical fault packets;

[0224] Construct a localization sub-model for the target segment based on all mapping sub-states;

[0225] The location sub-models for each power distribution section are set sequentially;

[0226] Generate a fault location model based on all location sub-models;

[0227] A diagnostic intelligent agent is constructed based on the fault diagnosis model and the fault location model.

[0228] According to the first concept of this application, multiple monitoring points are set based on the structural parameters of the distribution network, and the distribution network is divided into multiple distribution sections through all monitoring points. The operating status of each monitoring point is analyzed by F to quickly screen abnormal sections that may have fault risks. The fault diagnosis model is used to analyze the results of each abnormal section to quickly locate the fault section and improve the inspection efficiency of the distribution network.

[0229] According to the second concept of this application, by constructing a dual diagnostic model, on the one hand, the operation status of each monitoring point is analyzed to achieve the initial screening of abnormal areas, and on the other hand, a fault location model is constructed to achieve dual verification of the fault section and the abnormal status of the monitoring point, thereby improving the fault location accuracy of the distribution network.

[0230] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and substitutions can be made without departing from the technical principles of this application, and these improvements and substitutions should also be considered within the scope of protection of this application.

Claims

1. A method for locating and diagnosing fault sections in a distribution network based on artificial intelligence, characterized in that, include: Multiple monitoring points are set based on the equipment parameters of the power distribution network, and multiple power distribution sections are generated based on all monitoring points. A diagnostic agent is constructed based on all power distribution sections, and the diagnostic agent generates fault location results based on the acquired fault feedback packets. Obtain the diagnostic record package according to the preset update time node, and determine whether to generate an update instruction based on the diagnostic record package; This includes setting up multiple order allocation sections, including: Establish a power distribution section sequence A, A = (a1, a2, ..., a3) i …a n ), where a i Let be the i-th power distribution section; n is the number of power distribution sections.

2. The method for locating and diagnosing fault sections in a distribution network based on artificial intelligence as described in claim 1, characterized in that, Constructing a diagnostic agent includes: Establish a sequence of monitoring points B, B = (b1, b2, ..., bb2) i …b m ), where b i Let m be the i-th monitoring point; m is the number of monitoring points. Based on the monitoring point sequence B, bi is sequentially set as the target monitoring point; Generate correlation evaluation values ​​between target monitoring points and each power distribution section; A segment mapping table for target monitoring points is set based on all associated evaluation values; Construct an evaluation sub-model for the target monitoring points; A diagnostic sub-model for the target monitoring point is generated based on the segment mapping table and the evaluation sub-model. Diagnostic sub-models for each monitoring point are generated sequentially; A fault diagnosis model is generated based on all the diagnostic sub-models.

3. The method for locating and diagnosing fault sections in a distribution network based on artificial intelligence as described in claim 2, characterized in that, Building a diagnostic agent also includes: Based on the power distribution section sequence A, a is set sequentially. i For the target section; Obtain historical fault packets for the target segment; Generate a mapping evaluation value between the target section and each monitoring point; Establish a mapping evaluation value sequence C, C = (c1, c2, ..., c3) i …c m ), where c i is the mapping evaluation value between the target segment and the i-th monitoring point; m is the number of monitoring points; Preset mapping evaluation value threshold C1; If c i >C1, set the i-th monitoring point as the mapping monitoring point of the target segment; Obtain all mapped monitoring points in the target area; Set the mapping sub-state of each mapping monitoring point based on historical fault packets; Construct a localization sub-model for the target segment based on all mapping sub-states; The location sub-models for each power distribution section are set sequentially; Generate a fault location model based on all location sub-models; A diagnostic agent is constructed based on the fault diagnosis model and the fault location model.

4. The method for locating and diagnosing fault sections in a distribution network based on artificial intelligence as described in claim 3, characterized in that, Generate mapping evaluation values ​​between the target section and each monitoring point, including: Based on the monitoring point sequence B, bi is sequentially set as the monitoring point to be compared; Generate a mapping evaluation value c between the monitoring point to be compared and the target section; Where θ1 represents the number of fault records in the historical fault packets of the target segment; s i The first-level fluctuation value of the monitoring point to be compared in the i-th fault record; The mapping evaluation values ​​between the target section and each monitoring point are generated sequentially.

5. The method for locating and diagnosing fault sections in a distribution network based on artificial intelligence as described in claim 4, characterized in that, Generate fault location results, including: b is set sequentially according to the monitoring point sequence B. i These are monitoring points to be evaluated. Generate sub-feedback packets for the monitoring points to be evaluated based on the fault feedback packets; The diagnostic sub-model for the monitoring points to be evaluated is set as a primary diagnostic model; Generate the abnormal risk value f of the monitoring point to be evaluated based on the primary diagnostic model and sub-feedback package; Preset anomaly risk threshold F1; If f > F1, obtain all associated segments of the monitoring point to be evaluated according to the primary diagnostic model; Each associated segment of the monitoring point to be evaluated is designated as an abnormal segment; The abnormal risk values ​​for each monitoring point are generated sequentially. Establish an abnormal segment sequence A1 based on all abnormal risk values, where A1 = (a 11 a 12 …a 1i …a 1n1 ), where a 1i Let n1 be the i-th abnormal segment; n1 is the number of abnormal segments. Fault location results are generated based on all abnormal sections.

6. The method for locating and diagnosing fault sections in a distribution network based on artificial intelligence as described in claim 5, characterized in that, Generate the abnormal risk value f for the monitoring points to be evaluated, including: Where θ2 is the number of monitoring indicators for the monitoring points to be evaluated; β i v is the influence factor of the i-th monitoring indicator among the monitoring points to be evaluated; i It is the first-level reference value of the i-th monitoring indicator among the monitoring points to be evaluated, generated based on the sub-feedback package; v' i It is the standard reference value of the i-th monitoring indicator among the monitoring points to be evaluated, generated based on the primary diagnostic model.

7. The method for locating and diagnosing fault sections in a distribution network based on artificial intelligence as described in claim 6, characterized in that, Fault location results are generated based on all abnormal sections, including: Based on the abnormal segment sequence A1, a is set sequentially. 1i The target abnormal section; The target anomaly segment is defined as the target localization model; The matching evaluation value d of the target abnormal section is generated based on the target localization model and the fault feedback package; Consistently generate the matching evaluation values ​​for each abnormal segment; Establish a sequence of matching evaluation values ​​D, where D = (d1, d2, ..., dn). i …d n1 ), where d i is the matching evaluation value of the i-th abnormal segment; n1 is the number of abnormal segments; The preset matching evaluation threshold D1; Obtain the maximum value d in the sequence of matching evaluation values ​​D. max ; If d max >D1, generate first-level positioning instructions If d max <D1, set the maximum value d max The corresponding abnormal section is the target fault section; The first-level positioning instructions include: If di > d max， The i-th abnormal segment is set as the target fault segment.

8. The method for locating and diagnosing fault sections in a distribution network based on artificial intelligence as described in claim 7, characterized in that, Generate a matching evaluation value d for the target anomaly segment, including: Where θ3 is the number of mapping monitoring points in the target localization model; η i k represents the influence factor of the i-th mapping monitoring point in the target localization model. i To generate a similarity value for the i-th mapped monitoring point based on the fault feedback packet.

9. An artificial intelligence-based distribution network fault section location and diagnosis system, employing the artificial intelligence-based distribution network fault section location and diagnosis method as described in any one of claims 1-8, comprising: The central control unit is used to set multiple monitoring points according to the equipment parameters of the power distribution network; The central control unit is also used to generate multiple power distribution sections based on all monitoring points; The monitoring unit includes multiple monitoring sub-modules, which are located at various monitoring points. The monitoring unit is used to collect operational data from each monitoring point; The monitoring unit is also used to generate fault feedback packages; The central control unit includes: The first processing module is used to construct a diagnostic intelligent agent based on all power distribution sections. The diagnostic intelligent agent generates fault location results based on the acquired fault feedback packets. The second processing module is used to obtain the diagnostic record package according to the preset update time node, and determine whether to generate an update instruction based on the diagnostic record package. The third processing module is used to establish the power distribution section sequence A, A = (a1, a2, ..., a... i …a n ), where a i Let be the i-th power distribution section; n is the number of power distribution sections.

10. The artificial intelligence-based distribution network fault location and diagnosis system as described in claim 9, characterized in that, The first processing module is also used for: Establish a sequence of monitoring points B, B = (b1, b2, ..., bb2) i …b m ), where b i Let m be the i-th monitoring point; m is the number of monitoring points. Based on the monitoring point sequence B, bi is sequentially set as the target monitoring point; Generate correlation evaluation values ​​between target monitoring points and each power distribution section; A segment mapping table for target monitoring points is set based on all associated evaluation values; Construct an evaluation sub-model for the target monitoring points; A diagnostic sub-model for the target monitoring point is generated based on the segment mapping table and the evaluation sub-model. Diagnostic sub-models for each monitoring point are generated sequentially; Generate a fault diagnosis model based on all diagnostic sub-models; Based on the power distribution section sequence A, a is set sequentially. i For the target section; Obtain historical fault packets for the target segment; Generate a mapping evaluation value between the target section and each monitoring point; Establish a mapping evaluation value sequence C, C = (c1, c2, ..., c3) i …c m ), where c i is the mapping evaluation value between the target segment and the i-th monitoring point; m is the number of monitoring points; Preset mapping evaluation value threshold C1; If c i >C1, set the i-th monitoring point as the mapping monitoring point of the target segment; Obtain all mapped monitoring points in the target area; Set the mapping sub-state of each mapping monitoring point based on historical fault packets; Construct a localization sub-model for the target segment based on all mapping sub-states; The location sub-models for each power distribution section are set sequentially; Generate a fault location model based on all location sub-models; A diagnostic agent is constructed based on the fault diagnosis model and the fault location model.

Images