Multi-layer network information convergence analysis and collaborative filtering maintenance system and use method

Abstract

The present application relates to a kind of network data analysis technical field, it is a kind of multi-layer distribution network information convergence analysis and collaborative filtering overhaul system, method, comprising: service core control layer completes system scheduling and operation control;Resource control layer carries out resource management;Data acquisition transmission layer gathers distribution network equipment sensing data and power grid business resource data, and the data quality and transmission process are monitored;Abnormal analysis processing layer utilizes a variety of algorithms to identify and screen abnormal data in distribution network equipment sensing data and power grid business resource data, abnormal service research and judge according to abnormal data, and generate repair work order.The present application is based on the equipment infrared data of distribution network sensing layer, transformer temperature and humidity, cable voltage and current data and distribution network line, transformer load, voltage, current, substation information and other data are carried out hierarchical comprehensive calculation, solve the problem of inaccurate and not timely caused by single dimension of traditional distribution network information monitoring in abnormal information determination.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of power distribution network data analysis, and is a multi-layer power distribution network information convergence analysis and collaborative filtering maintenance system and a use method. BACKGROUND

[0002] In recent years, the scale of power distribution network data is increasing, and the requirements for the operation state of power distribution network equipment and power supply quality are also increasing. However, the increasing scale of power distribution network data leads to more frequent changes, and therefore, abnormal monitoring needs to be set. Traditional abnormal monitoring and processing methods are mostly separated, offline processed, and operated in processes such as maintenance and data maintenance. The abnormal monitoring process is complicated and cannot meet the requirements of data real-time and timeliness. In the maintenance process, a work order needs to be manually initiated, business classification needs to be manually performed, and a maintenance plan needs to be manually declared, which is low in work efficiency, inaccurate in plan declaration, and not timely in processing of change information. At the same time, daily work tickets, operation tickets, inspection, defect records, and maintenance plan input are all in a production environment, and therefore, the work efficiency of power distribution network abnormal monitoring and maintenance processing is low. SUMMARY

[0003] The application provides a multi-layer power distribution network information convergence analysis and collaborative filtering maintenance system and a use method, which overcomes the shortcomings of the prior art, and effectively solves the problems of low efficiency, poor real-time performance, manual generation of work orders, and errors in existing power distribution network abnormal monitoring.

[0004] One of the technical solutions of the application is realized through the following measures: a multi-layer power distribution network information convergence analysis and collaborative filtering maintenance system, comprising:

[0005] a service core control layer, which completes system scheduling and operation control, realizes power distribution network information convergence analysis and collaborative filtering maintenance, and is connected with the data acquisition transmission layer, the abnormal analysis processing layer, and the big data analysis display layer;

[0006] a resource control layer, which performs engine resource management, user permission control, and work flow management, and is connected with the data acquisition transmission layer, the abnormal analysis processing layer, and the big data analysis display layer;

[0007] a data acquisition transmission layer, which collects power distribution network equipment sensing data and power grid business resource data, and monitors data quality and a transmission process;

[0008] an abnormal analysis processing layer, which identifies and screens abnormal data in the power distribution network equipment sensing data and the power grid business resource data by using a plurality of algorithms, performs abnormal business research and judgment according to the abnormal data, and generates a maintenance work order;

[0009] a big data analysis display layer, which performs big data analysis and visual display on data of the data acquisition transmission layer and the abnormal analysis processing layer.

[0010] The following is a further optimization or / and improvement of the above-mentioned technical solutions of the application:

[0011] The data collection transmission layer comprises a data collection layer and a data monitoring layer.

[0012] The data collection layer collects and stores power distribution equipment sensing data and power grid business resource data.

[0013] The data monitoring layer performs quality analysis and data transmission monitoring on the data, and displays and marks the abnormality, wherein the quality analysis comprises data attribute extraction, format conversion, dimension definition and rule checking.

[0014] The data collection layer comprises an integrated database, a power distribution equipment sensing data collection unit and a power grid business resource data collection unit.

[0015] The power distribution equipment sensing data collection unit comprises a front-end sensing module, a sensing data buffer module and a sensing data conversion processing module, the front-end sensing module collects the front-end power distribution equipment sensing data according to the collection instruction of the service core control layer, and buffers the power distribution equipment sensing data to the sensing data buffer module, the sensing data conversion processing module extracts the power distribution equipment sensing data in the sensing data buffer module, and stores the power distribution equipment sensing data after marking classification, data protocol conversion and A / D conversion to the integrated database.

[0016] The power grid business resource data collection unit comprises a front-end collection module, a resource data buffer module and a resource data conversion processing module, the front-end collection module extracts and re-operates the required power grid business resource data in the power grid business resource library, and buffers the operated power grid business resource data to the resource data buffer module, the resource data conversion processing module extracts the power grid business resource data in the resource data buffer module, and stores the power grid business resource data after marking classification, data processing and data protocol conversion to the integrated database.

[0017] The data monitoring layer comprises a quality monitoring unit and a transmission monitoring unit.

[0018] The quality monitoring unit comprises a data quality monitoring analysis module, a benchmark database, a data dimension definition statistical module, a data detection module, a checking module and an analysis result display module, the data quality monitoring analysis module extracts the data collected by the data collection layer and the benchmark data in the benchmark database according to the instruction, the data detection module extracts the data attribute, the data dimension definition statistical module performs format conversion and dimension definition on the data, the checking module performs rule checking on the data processed by the data dimension definition statistical module, and the analysis result display module displays the rule checking result.

[0019] The transmission monitoring unit comprises a token bucket data transmission monitoring module, a double-rate three-color marking module, an information rate calculation module, a message analysis and judgment module, and a token bucket marking cache module. The token bucket data transmission monitoring module controls the double-rate three-color marking module to perform double-rate three-color marking according to instructions. The information rate calculation module calculates the information rate of data. The message analysis and judgment module compares the marking information and the calculation result, judges and separates abnormal transmission data, and caches the abnormal transmission data in the token bucket marking cache module.

[0020] The abnormal analysis processing layer comprises an abnormal data operation layer, an abnormal service judgment layer, and a work order triggering layer.

[0021] The abnormal data operation layer pre-processes the distribution network equipment sensing data and power grid service resource data, performs data fusion after the pre-processing, and identifies and screens abnormal data using various algorithms after the fusion.

[0022] The abnormal service judgment layer judges abnormal services according to abnormal data, and marks abnormal services corresponding to the abnormal data.

[0023] The work order triggering layer determines an abnormal service maintenance scheme according to the marked abnormal services, and generates and pushes a maintenance work order after verification.

[0024] The abnormal data operation layer comprises:

[0025] The sensing data preprocessing module and the power grid service resource data preprocessing module respectively perform data cleaning on the distribution network equipment sensing data and the power grid service resource data. The data rectification operation module performs secondary data cleaning and data structure verification, and transmits the processed data to the data fusion module through the BUS bus for data fusion. The data prediction module defines and converts the data protocol of the fused data, and pushes the converted data matrix arrangement to the secondary data analysis module after completion. The data logic correlation operation module performs matrix data operation, the data marking module performs service data marking, and the data attribute target identification module performs fused data identification and screening after completion. The data is pushed to the fusion database.

[0026] The data analysis logic control unit extracts fused data, and analyzes the data using an OCSVM abnormal data detection control module and a Kmeans abnormal data detection control module.

[0027] The data output by the OCSVM abnormal data detection control module is separated in density by the density separation module through Gaussian distribution statistics and calculation of data distribution variance, classified by the data high-dimensional detection after completion, and the abnormal data is dispersed. The abnormal data is marked by the abnormal data marking module, and pushed to the multi-dimensional abnormal data marking module.

[0028] The data output by the Kmeans abnormal data detection control module is pushed to the clustering operation module for clustering separation, the convex set intersection operation is completed through the convex data set logical operation module, the set of n-dimensional positive semi-definite matrices is formed, the convex data set convergence is performed through the KMeans convergence algorithm, the converged data set is resolved, and is pushed to the multi-dimensional abnormal data marking module;

[0029] The multi-dimensional abnormal data marking module performs intersection operation to form a multi-dimensional abnormal data set, and pushes the multi-dimensional abnormal data set to the multi-dimensional abnormal data logical judgment module for data format and protocol operation, and pushes the message to the service core control layer.

[0030] The abnormal service research and judgment layer, the abnormal data trigger module calls the Shark abnormal service shunt module to separate the abnormal data logical attributes, and pushes the abnormal data logical attributes to the abnormal service analysis operation module for service logic classification and definition. The thread operation module performs service classification and flow transfer, and according to different service logic attributes, flows to the fault repair plan logical control module, the active repair logical control module, and the planned power outage logical control module. The abnormal service marking display module is called to complete the marking of the corresponding abnormal service attributes, and returns to the service core control layer.

[0031] Or / and,

[0032] The work order trigger layer, the work order trigger unit calls the Jbpm work order scheduling control module, pre-generates a service work order tag according to an abnormal service maintenance scheme, and the work order logical control unit calls the corresponding fault repair work order control module, active repair work order control module, and planned power outage work order control module according to the work order tag. The work order data and the work order tag are pushed to the work order data marking unit for logical linkage. The work order data verification module verifies the work order data and the work order tag with the abnormal data. After verification succeeds, the work order data and the work order tag are pushed to the work order data verification cache. The work order data verification logical research and judgment module performs logical verification operation on the service tag and the work order tag, generates a maintenance work order after completion, and returns to the service core control layer.

[0033] The above-mentioned resource control layer includes a Calibrated boosted trees cloud computing module, a logical operation module, a virtualization resource management module, a user permission module, and a workflow management module. The Calibrated boosted trees cloud computing module calls the logical operation module to control the virtualization resource management module, the user permission module, and the workflow management module respectively to perform engine resource management, user permission control, and workflow management.

[0034] Or / and,

[0035] The big data analysis display layer, the data preprocessing module pushes the marked completed abnormal business attribute and data to the data preprocessing module for preprocessing, pushes to the metadata management module and data image conversion module after completion, respectively carries out data definition and data image conversion, the data management module pushes data to the FFT data operation module and EMTraining data operation module respectively for data analysis, pushes to the BI interactive module after completion, the data image conversion module pushes the image data to the data visualization analysis logic control unit for image data and data attribute logical correlation, pushes to the BI interactive module after completion, the BI interactive module carries out structured and unstructured data application classification and marking, pushes to the visualization display control module after completion, carries out data and big data graphics display through the perception layer data visualization model and power grid business resource data visualization model;

[0036] Or / and,

[0037] The service core control layer, the service center total control unit completes data bidirectional transmission and instruction control through the FCFS scheduling unit, and stores data to the center database, and the FCFS scheduling unit calls the corresponding logic control unit of each layer to realize instruction sending, running control and data transmission between each layer and the service core control layer.

[0038] The second technical scheme of the application is realized by the following measures: a multi-layer distribution network information convergence analysis and collaborative filtering maintenance method, comprising:

[0039] The service core control layer controls the resource control layer, the data acquisition and transmission layer, the abnormal analysis processing layer and the big data analysis display layer to initialize;

[0040] In response to the initialization success instruction, the service core control layer calls the resource control layer to perform engine resource management, user permission control and workflow management;

[0041] In response to the resource control layer working normally, the service core control layer calls the data acquisition and transmission layer, acquires distribution network equipment sensing data and power grid business resource data, and monitors data quality and transmission process;

[0042] The service core control layer calls the abnormal analysis processing layer to identify and filter abnormal data in the distribution network equipment sensing data and the power grid business resource data by using multiple algorithms, and carries out abnormal business research and judgment according to the abnormal data;

[0043] The service core control layer calls the big data analysis display layer to generate an abnormal business maintenance scheme corresponding to the abnormal business;

[0044] The service core control layer calls the abnormal analysis processing layer to generate a maintenance work order based on the abnormal business maintenance scheme;

[0045] The service core control layer calls a big data analysis display layer to perform visual display of data.

[0046] The beneficial effects of the present application include:

[0047] The present application performs hierarchical comprehensive calculation based on infrared data of devices of a distribution network sensing layer, transformer temperature and humidity, cable voltage and current data, and distribution network line, transformer load, voltage, current, and substation information, etc., to solve the problem of inaccurate and untimely abnormal information determination caused by single dimension of traditional distribution network information monitoring. And the service core control layer, resource control layer, data acquisition and transmission layer, abnormal analysis and processing layer, and big data analysis display layer are utilized to realize automatic processing of data acquisition, processing, research and judgment, work order, and display in all directions, and to efficiently and quickly realize distribution network information convergence analysis and collaborative filtering repair, compared with the existing single abnormal analysis, abnormal business without research and judgment, and manual initiation of repair work order mode, the efficiency and accuracy of abnormal analysis and work order generation are effectively improved.

[0048] The present application technically fuses sensing layer data and power grid information system business data through a data fusion process, compared with the problem of single data acquisition and analysis process in traditional business, data islands between data, and different format data unable to be linked for analysis.

[0049] In the abnormal business research and judgment of the present application, abnormal data and business scenarios are automatically associated based on abnormal business through a shark algorithm, to form abnormal business research and judgment, and to solve the problem of inaccurate information and positioning error of artificially defined abnormal business in traditional business.

[0050] In the present application, the data quality and data transmission in the data acquisition process and data operation and analysis process are monitored in real time and in the whole process through the data monitoring layer, to guarantee the accuracy, integrity and timeliness of data from the source end.

[0051] In the present application, the FCFS scheduling algorithm is utilized to complete scheduling of instructions and data, to improve the independent running property of overall engine data and transmission, and to improve the overall work efficiency.

[0052] In the present application, the accuracy of work order information is verified when generating a work order, to effectively avoid the delay of repair work caused by work order information error in traditional repair business or re-declaration plan, and to prevent further expansion of losses caused by untimely repair. BRIEF DESCRIPTION OF DRAWINGS

[0053] ATTACHED Figure 1 The figure is a system structure schematic diagram of the present application.

[0054] ATTACHED Figure 2 The figure is a structure schematic diagram of the service core control layer in the present application.

[0055] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 3 Figure 1 is a structural schematic diagram of a resource control layer in the application.

[0056] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 4 Figure 1 is a structural schematic diagram of a resource control layer in the application.

[0057] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 5 Figure 1 is a structural schematic diagram of a resource control layer in the application.

[0058] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 6 Figure 1 is a structural schematic diagram of a resource control layer in the application.

[0059] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 7 Figure 1 is a structural schematic diagram of a resource control layer in the application.

[0060] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 8 Figure 1 is a structural schematic diagram of a resource control layer in the application.

[0061] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 9 Figure 1 is a structural schematic diagram of a resource control layer in the application.

[0062] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 10 Figure 1 is a structural schematic diagram of a resource control layer in the application.

[0063] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 11 Figure 1 is a structural schematic diagram of a resource control layer in the application.

[0064] Figure 1 is a structural schematic diagram of a resource control layer in the application. Figure 12 Figure 1 is a structural schematic diagram of a resource control layer in the application. DETAILED DESCRIPTION

[0065] The application is not limited by the following examples, and the specific implementation can be determined according to the technical solution of the application and the actual situation.

[0066] The application will be further described below in combination with examples and drawings:

[0067] Example 1: as shown in the accompanying drawings, the application embodiment discloses a multi-layer distribution network information convergence analysis and collaborative filtering maintenance system, which comprises: Figure 1 (1) a service core control layer, which completes system scheduling and operation control and realizes distribution network information convergence analysis and collaborative filtering maintenance.

[0068] Specifically, as shown in the accompanying drawings, the application embodiment discloses a multi-layer distribution network information convergence analysis and collaborative filtering maintenance system, which comprises:

[0069] Figure 2 ​As shown, the service center control unit in the service core control layer completes data bidirectional transmission and instruction control through the FCFS scheduling unit, and stores the data to the center database, and the FCFS scheduling unit calls the corresponding logical control unit of each layer to realize instruction sending, running control and data transmission between each layer and the service core control layer.

[0070] It should be noted that the corresponding logical control unit of each layer of the service core control layer includes:

[0071] The resource control unit is used for corresponding to the resource control layer, and the FCFS scheduling unit calls the resource control unit according to the instruction of the service center control unit to complete the internal running control of the resource control layer.

[0072] The big data analysis control unit, the FCFS scheduling unit calls the big data analysis control unit according to the instruction of the service center control unit to complete the internal running of the big data analysis display layer.

[0073] The data acquisition logical control unit, the XPDL data interaction unit and the data monitoring logical control unit, the FCFS scheduling unit calls the data acquisition logical control unit, the XPDL data interaction unit and the data monitoring logical control unit according to the instruction of the service center control unit to complete the internal running control of the data acquisition transmission layer.

[0074] The data operation logical control unit, the business research and judgment logical control unit and the work order logical control unit, the FCFS scheduling unit calls the data operation logical control unit, the business research and judgment logical control unit and the work order logical control unit according to the instruction of the service center control unit to complete the internal running control of the abnormal analysis processing layer.

[0075] (2) Resource control layer, engine resource management, user permission control and workflow management.

[0076] Specifically, as shown in the accompanying Figure 3 As shown, the resource control layer includes a Calibrated boosted trees cloud computing module, a logical operation module, a virtualization resource management module, a user permission module and a workflow management module, the Calibrated boosted trees cloud computing module calls the logical operation module to control the virtualization resource management module, the user permission module and the workflow management module respectively, and performs engine resource management, user permission control and workflow management respectively.

[0077] (3) Data acquisition transmission layer, acquires power distribution network equipment sensing data and power grid business resource data, and monitors data quality and transmission process.

[0078] Specifically, as shown in the accompanying Figure 4 As shown, the data acquisition transmission layer includes a data acquisition layer and a data monitoring layer.

[0079] (1) Data acquisition layer, collecting and storing power distribution equipment sensing data and power grid business resource data.

[0080] Among them, as shown in the accompanying Figure 5 The data acquisition layer includes an integrated database, a power distribution equipment sensing data acquisition unit, and a power grid business resource data acquisition unit.

[0081] The power distribution equipment sensing data acquisition unit includes a front-end sensing module, a sensing data cache module, and a sensing data conversion processing module. The front-end sensing module collects front-end power distribution equipment sensing data (data in the form of serial code) according to the collection instructions of the service core control layer, pushes the collected serial code into the sensing data cache module for caching, and the sensing data conversion processing module includes a sensing data marking module, a sensing data marking module, and a sensing collection logic operation module. The sensing data marking module marks and classifies the power distribution equipment sensing data, and after completion, pushes it to the sensing data protocol conversion module, converts the collected analog data into digital data, and after completion, pushes the digital data to the sensing collection logic operation unit for A / D conversion, and then stores it in the integrated database. It should be further explained that the front-end sensing module can include a power distribution equipment infrared data acquisition module, a transformer temperature and humidity data acquisition module, a cable voltage data acquisition module, and a cable current data acquisition module. The collected power distribution equipment sensing data includes power distribution equipment infrared data, transformer temperature and humidity data, cable voltage data, and cable current data. Here, the number of sensing data cache modules is the same as and corresponds one-to-one to the collection modules in the front-end sensing module.

[0082] The power grid business resource data acquisition unit includes a front-end acquisition module, a resource data caching module, and a resource data conversion and processing module. The front-end acquisition module includes load rate logic calculation modules, overload logic calculation modules, undervoltage logic calculation modules, three-phase imbalance logic calculation modules, and openable capacity logic calculation modules. The resource data caching module corresponds to these modules. Specifically, the load rate logic calculation modules, overload logic calculation modules, undervoltage logic calculation modules, three-phase imbalance logic calculation modules, and openable capacity logic calculation modules extract relevant resources from the power grid business resource library through an API interface module. For example, the load rate logic calculation module calculates the load rate based on different dimensions of load data using logical algorithms such as time, transformer area, and equipment type, and pushes the results to the load rate data cache. The overload logic calculation module extracts data such as line load, equipment type, and transformer area, calculates the overload value using different dimensions of logical algorithms, and pushes the results to the overload data cache. The undervoltage logic calculation module converts line voltage, bus balance, load, and transformer area information into data based on different dimensions of logical algorithms. The algorithm calculates low voltage and pushes the results to the low voltage data cache. The three-phase imbalance logic operation module calculates three-phase voltage based on phase voltage, current, load, and transformer area information using different logic algorithms and pushes the results to the three-phase imbalance voltage data cache. The openable capacity logic operation module calculates openable capacity data based on line rated capacity, load factor, power load, and transformer area information using different logic algorithms and pushes the results to the openable capacity data cache. After resource extraction, the data is pushed to the resource data conversion and processing unit. The resource data conversion and processing unit includes a power grid business resource data marking module, a power grid business resource data protocol conversion module, and a power grid business resource data logic operation module. The power grid business resource data marking module marks and classifies the corresponding business data and pushes it to the power grid business resource data protocol conversion module. After data protocol conversion, the data is pushed to the power grid business resource data logic operation module to calculate load rate, overload, low voltage, three-phase imbalance, and openable capacity values ​​and pushes the results to the integrated database.

[0083] (2) Data monitoring layer, which performs quality analysis and data transmission monitoring on the data, and displays and marks anomalies. The quality analysis includes data attribute extraction, format conversion, dimension definition and rule verification.

[0084] For details, see attached. Figure 6 As shown, the data monitoring layer includes a quality monitoring unit and a transmission monitoring unit;

[0085] The quality monitoring unit comprises a data quality monitoring analysis module, a benchmark database, a data dimension definition statistics module, a data detection module, a verification module and an analysis result display module. The data quality monitoring analysis module extracts data collected by the data collection layer and benchmark data in the benchmark database according to an instruction. The data detection module extracts data properties, including the number of non-empty values, the number of non-repeated values, the maximum value, the minimum value, the number of top 5 values and the like. The data dimension definition statistics module performs format conversion and dimension definition on the data. The verification module performs rule verification on the data processed by the data dimension definition statistics module, including the completeness, correctness, currentness and consistency of the data. The analysis result display module displays the rule verification result.

[0086] It should be noted that the quality monitoring unit can also monitor the abnormal data identified by the abnormal analysis processing layer, and the monitoring method is the same.

[0087] The transmission monitoring unit comprises a token bucket data transmission monitoring module, a dual-rate three-color marking module, an information rate calculation module, a message analysis and judgment module and a token bucket marking cache module. The token bucket data transmission monitoring module controls the dual-rate three-color marking module to perform dual-rate three-color marking according to an instruction. The information rate calculation module calculates the information rate of the data. The message analysis and judgment module compares the marking information and the calculation result, judges and separates abnormal transmission data, and caches the abnormal transmission data in the token bucket marking cache module.

[0088] It should be noted that the information rate calculation module comprises a CIR operation module, a CBS operation module, a PIR operation module and a PBS operation module.

[0089] (4) Abnormal analysis processing layer, using multiple algorithms to identify and filter abnormal data in the power distribution equipment sensing data and power grid business resource data, according to the abnormal data to carry out abnormal business research and judgment, and generate repair work order.

[0090] Specifically, as shown in FIG. 4, the abnormal analysis processing layer comprises an abnormal data operation layer, an abnormal business research and judgment layer and a work order triggering layer. Figure 7

[0091] The abnormal data operation layer pre-processes the power distribution equipment sensing data and power grid business resource data, performs data fusion after the pre-processing, and identifies and filters abnormal data using multiple algorithms after the fusion.

[0092] The abnormal business research and judgment layer performs abnormal business research and judgment according to the abnormal data, and marks the abnormal business corresponding to the abnormal data.

[0093] The work order triggering layer determines an abnormal business repair scheme according to the marked abnormal business, and generates and pushes a repair work order after verification.

[0094] ​wherein, as shown in the accompanying drawings Figure 8 The abnormal data operation layer includes:

[0095] The perception data preprocessing module and the power grid service resource data preprocessing module respectively perform data cleaning (the data cleaning uses existing conventional data preprocessing methods, such as missing value, outlier, inconsistency, etc.) on the distribution network equipment perception data and the power grid service resource data. The data rectification operation module performs secondary data cleaning and data structure verification, and transmits the processed data to the data fusion module through the BUS bus for data fusion. The data prediction module defines and converts the data protocol of the fused data, and after completion, pushes to the secondary data analysis module for matrix arrangement of the converted data. The data logic correlation operation module performs matrix data operation, the data labeling module performs service data labeling, and after completion, pushes to the data attribute target identification module for identification and screening of the fused data, and after completion, pushes to the fusion database.

[0096] The data analysis logic control unit extracts the fused data, and analyzes by using the OCSVM abnormal data detection control module and the Kmeans abnormal data detection control module respectively.

[0097] Specifically, the data output by the OCSVM abnormal data detection control module is separated by the density separation module through Gaussian distribution statistics and calculation of data distribution variance, and after completion, the data set is classified by the data high-dimensional detection. Through different parameter adjustment of the base detector, based on sample feature sampling, the calculation of abnormal vector is formed, the normalized value is converted, the abnormal data is discretized, and the discretized abnormal data is labeled by the abnormal data labeling module and pushed to the multi-dimensional abnormal data labeling module. The data output by the Kmeans abnormal data detection control module is pushed to the clustering operation module for clustering separation, the convex set intersection operation is completed by the convex data set logic operation module, the n-dimensional positive semi-definite matrix set is formed, the convex data set is converged by the KMeans convergence algorithm, the converged data set is resolved, and pushed to the multi-dimensional abnormal data labeling module.

[0098] The multi-dimensional abnormal data labeling module performs intersection operation to form a multi-dimensional abnormal data set, which is pushed to the multi-dimensional abnormal data logic judgment module for data format and protocol operation, and is pushed to the data operation logic control module of the service core control layer in the form of a message.

[0099] wherein, as shown in the accompanying drawings Figure 9As shown, the abnormal service research and judgment layer includes an abnormal data trigger module calling a Shark abnormal service shunting module to separate abnormal data logical attributes, and push them to an abnormal service analysis and operation module for service logic classification and definition. After completion, a thread operation module performs service classification and flow transfer, and according to different service logic attributes, flows to a fault repair plan logic control module, an active repair logic control module, and a planned power outage logic control module for related business logic calculation. After completion, an abnormal service marking display module is called to complete the marking of corresponding abnormal service attributes, and returns to the service core control layer.

[0100] As shown in the accompanying drawings, Figure 10 As shown, the work order trigger layer includes a work order trigger module calling a Jbpm work order scheduling control module to pre-generate a business work order tag according to an abnormal service maintenance scheme. A work order logic control module calls corresponding fault repair work order control module, active repair work order control module, and planned power outage work order control module according to the work order tag to review corresponding work order data. The work order data and the work order tag are pushed to a work order data marking module for logical linkage. A work order data verification module verifies the work order data and the work order tag with abnormal data extracted by the service center master control unit. After successful verification, the work order data and the work order tag are pushed to a work order data verification cache. A work order data verification logic research and judgment module performs logical verification and operation on the business tag and the work order tag. After completion, a maintenance work order is generated and returned to the work order logic control unit in the service core control layer.

[0101] (Five) Big data analysis and display layer, which performs big data analysis and visual display on the data of the data collection and transmission layer and the abnormal analysis and processing layer.

[0102] As shown in the accompanying drawings, Figure 11 As shown in the accompanying drawings,

[0103] The application discloses a kind of multi-layer distribution network information convergence analysis and collaborative filtering maintenance system, based on the equipment infrared data of distribution network sensing layer, transformer temperature and humidity, cable voltage and current data and distribution network line, transformer load, voltage, current, substation information etc.Data are hierarchically calculated comprehensively, solve the problem of inaccurate and untimely abnormal information determination caused by single dimension of traditional distribution network information monitoring.Further, the application uses service core control layer, resource control layer, data acquisition transmission layer, abnormal analysis processing layer and big data analysis display layer to realize data acquisition, processing, research and judgment, work order, display omnidirectional automatic processing, efficiently and quickly realizes distribution network information convergence analysis and collaborative filtering maintenance, compared with the existing single abnormal analysis, abnormal service without research and judgment, maintenance work order manual initiation mode, effectively improve the efficiency and accuracy of abnormal analysis and work order generation.

[0104] Embodiment 2: as shown in the attached Figure 12 The application discloses a kind of multi-layer distribution network information convergence analysis and collaborative filtering maintenance system's use method, including:

[0105] Step S101, service core control layer controls resource control layer, data acquisition transmission layer, abnormal analysis processing layer and big data analysis display layer initialization.

[0106] Specifically, service center general control unit sends instruction to FCFS scheduling unit, and FCFS scheduling unit sends initialization instruction to resource control layer, data acquisition transmission layer, abnormal analysis processing layer and big data analysis display layer, and resource control layer, data acquisition transmission layer, abnormal analysis processing layer and big data analysis display layer execute initialization instruction.

[0107] Step S102, in response to initialization success instruction, service core control layer calls resource control layer to carry out engine resource management, user permission control and workflow management.

[0108] Specifically, service center general control unit can judge whether initialization is completed according to initialization execution result feedback information, in response to no, then re-initialization, in response to yes, then service center control unit calls resource control unit through FCFS scheduling unit, controls Calibrated boosted trees cloud computing module, logic operation module, virtualization resource management module, user permission module, workflow management module to carry out virtualization resource management, user permission management and workflow management.

[0109] Step S103, in response to resource control layer working normally, service core control layer calls data acquisition transmission layer, acquires distribution network equipment sensing data and power grid business resource data, and monitors data quality and transmission process.

[0110] Specifically, the service center total control unit can determine whether the resource management is normal according to the feedback information, in response to no, recheck the resources, in response to yes, the service center control unit calls the data collection logic control unit through the FCFS scheduling unit, and controls the data collection transmission layer to collect and monitor data.

[0111] Specific data collection is as follows:

[0112] Step S311, the data collection logic unit sends instructions through the XPDL data interaction module;

[0113] Step S312, the distribution network equipment infrared data collection module, the transformer temperature and humidity data collection module, the cable voltage data collection module and the cable current data collection module collect distribution network equipment sensing data including distribution network equipment infrared data, transformer temperature and humidity data, cable voltage data and cable current data.

[0114] Step S313, the collected distribution network equipment sensing data is pushed into the sensing data cache module for caching, the sensing data marking module marks and classifies the distribution network equipment sensing data, and after completion, pushes the sensing data to the sensing data protocol conversion module, converts the collected analog data into digital data, after completion, pushes the digital data to the sensing collection logic operation unit for A / D conversion, and stores the digital data in the integrated database;

[0115] Step S314, the load rate logic operation module, the overload logic operation module, the low voltage logic operation module, the three-phase imbalance logic operation module and the openable capacity logic operation module extract load rate, overload rate, low voltage, three-phase imbalance data and openable capacity related data in the power grid business resource library through the API interface module.

[0116] Step S315, the power grid business resource data marking module marks and classifies the corresponding business data, after completion, pushes the data to the power grid business resource data protocol conversion module, after completion of data protocol conversion, pushes the data to the power grid business resource data logic operation unit for load rate, overload, low voltage, three-phase imbalance and openable capacity value operation, and after completion, pushes the data to the integrated database.

[0117] Specific data monitoring is as follows:

[0118] Step S321, the service center control unit calls the data monitoring logic control unit through the FCFS scheduling unit to monitor the overall data quality and transmission process.

[0119] Step S322, the data quality monitoring analysis module extracts the data collected by the data collection layer and the reference data in the reference database according to the instruction, the data detection module extracts the data attributes, including the number of non-empty, the number of non-repeated values, the maximum value, the minimum value, the number of top5 values, etc., the data dimension definition statistics module performs format conversion and dimension definition on the data, the verification module performs rule verification on the data processed by the data dimension definition statistics module, including the integrity, correctness, currentness and consistency of the data, and the analysis result display module displays the rule verification result.

[0120] Step S323, the token bucket data transmission monitoring module controls the double-rate three-color marking module to perform double-rate three-color marking according to the instruction, the information rate calculation module calculates the information rate of the data, the message analysis and judgment module compares the marking information and the calculation result, judges and separates the abnormal transmission data, and buffers the abnormal transmission data in the token bucket marking buffer module.

[0121] Step S324, it is judged whether the data transmission monitoring is abnormal, if abnormal, the cycle monitoring is performed, if normal, the distribution network equipment sensing data and the power grid business resource data are pushed to the service center master control unit through the XPDL data interaction module.

[0122] Step S104, the service core control layer calls the abnormal analysis processing layer to identify and filter the abnormal data in the distribution network equipment sensing data and the power grid business resource data by using multiple algorithms, and performs abnormal business analysis according to the abnormal data.

[0123] Specifically, the data is pushed to the data logic operation unit by the service center master control unit, the abnormal data operation layer identifies and filters the abnormal data, that is, carries out data operation, prediction, secondary analysis, logical correlation, marking and attribute identification, completes the fusion of the sensing layer and the business resource data, and the data analysis logic control unit performs two-way abnormal monitoring on the fused data, one is to carry out OCSVM abnormal monitoring, and the abnormal data is marked after data density separation and high-dimensional detection, the other is to carry out Kmeans abnormal monitoring, and the hand chain data set is formed after cluster operation and convex data set operation. The two abnormal detection results are subjected to multi-dimensional data intersection logical operation, and the abnormal data is subjected to final logical judgment.

[0124] Specifically, the service center general control unit calls the service research and judgment logic control unit by using the FCFS scheduling unit, controls the abnormal service research and judgment layer to classify the abnormal data, classifies the abnormal service by the shark abnormal shunt module, carries out the instruction control of the scene service research and judgment by the thread operation unit, carries out the fault repair plan, active repair, planned power outage, completes the abnormal service scene research and judgment, marks and displays the abnormal data and abnormal service, and returns the result to the service research and judgment logic control unit.

[0125] In step S105, the service core control layer calls the big data analysis display layer to generate an abnormal service maintenance scheme corresponding to the abnormal service.

[0126] Specifically, the data management module in the big data analysis display layer can prestore abnormal service maintenance schemes corresponding to various abnormal services. After receiving the abnormal service, the corresponding abnormal service maintenance scheme can be directly matched and found.

[0127] In step S106, the service core control layer calls the abnormal analysis processing layer to generate a maintenance work order based on the abnormal service maintenance scheme.

[0128] Specifically, the service center control layer calls the work order logic control unit by using the FCFS scheduling unit, controls the work order trigger layer, that is, the work order logic control unit controls the work order trigger module, generates the corresponding fault repair, active repair, and planned power outage work order information based on the abnormal service maintenance scheme by using the jbpm work order scheduling control module, and carries out data verification according to the abnormal data and abnormal service to determine whether it is abnormal. If it is abnormal, the previous step is performed. If it is normal, a service abnormal maintenance work order is generated.

[0129] In step S107, the service core control layer calls the big data analysis display layer to perform visual display of data.

[0130] Specifically, the service center control layer calls the big data analysis control unit by using the FCFS scheduling unit, the service center general control unit pushes data to the big data analysis control module for data display, that is, the data preprocessing module pushes the marked abnormal business attributes and data to the data preprocessing module for preprocessing, pushes to the metadata management module and data image conversion module after completion, respectively performs data definition and data image conversion, the data management module pushes data to the FFT data operation module and the EM Training data operation module for data analysis, pushes to the BI interaction module after completion, the data image conversion module pushes image data to the data visualization analysis logic control unit for logical association of image data and data attributes, pushes to the BI interaction module after completion, the BI interaction module performs structured and unstructured data application classification and marking, pushes to the visualization display control module after completion, and performs data and big data graphic display through the perception layer data visualization model and the power grid business resource data visualization model.

[0131] Embodiment 3: The embodiment of the application discloses an electronic device, comprising a processor and a memory, the memory stores a computer program, the computer program is loaded and executed by the processor to realize the use method of the multi-layer network information convergence analysis and collaborative filtering maintenance system.

[0132] The processor can be a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an ASIC, an FPGA, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logic blocks, modules, and circuits described in combination with the disclosure. It can also be a combination of computing functions, such as one or more microprocessor combinations, DSP and microprocessor combinations, etc. The memory can include, but is not limited to, a U disk, a read-only memory, a mobile hard disk, a magnetic disk or an optical disk, and various computer program storage media.

[0133] Those skilled in the art should understand that the embodiments of the application can be provided as a method, a system, or a computer program product. Therefore, the application can be in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the application can be in the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code. The solutions in the embodiments of the application can be implemented in various computer languages, such as object-oriented programming languages Java and interpreted scripting languages JavaScript.

[0134] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0135] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 Figure 1 The function specified in one or more boxes.

[0136] The above technical features constitute the preferred embodiment of the present invention, which has strong adaptability and optimal implementation effect. Unnecessary technical features can be added or removed according to actual needs to meet the requirements of different situations.

Claims

1. A multi-layer network information convergence analysis and collaborative filtering maintenance system, characterized in that, Comprise: The service core control layer, complete system scheduling and operation control, realize distribution network information convergence analysis and collaborative filtering repair; Resource control layer, engine resource management, user permission control and workflow management; Data acquisition and transmission layer, collect distribution network equipment sensing data and power grid business resource data, and monitor data quality and transmission process; Abnormal analysis and processing layer, use multiple algorithms to identify and filter abnormal data in distribution network equipment sensing data and power grid business resource data, analyze abnormal business according to abnormal data, and generate repair work order; Big data analysis and display layer, big data analysis and visualization display of data collected by data acquisition and transmission layer and abnormal analysis and processing layer; Among them, the abnormal analysis and processing layer includes abnormal data operation layer, abnormal business analysis layer and work order triggering layer; Abnormal data operation layer, pre-process distribution network equipment sensing data and power grid business resource data, data fusion after pre-processing, and use multiple algorithms to identify and filter abnormal data after fusion; Abnormal business analysis layer, abnormal business analysis according to abnormal data, and marking abnormal data corresponding to abnormal business; Work order triggering layer, determine abnormal business repair scheme according to marked abnormal business, and generate and push repair work order after verification; Among them, the abnormal data operation layer includes: Sensing data preprocessing module and power grid business resource data preprocessing module respectively preprocess distribution network equipment sensing data and power grid business resource data, data rectification operation module carries out secondary data cleaning and data structure verification, and transmits the processed data to data fusion module through BUS bus for data fusion, data prediction module defines and converts the data protocol of the fused data, and pushes it to the secondary data analysis module for data matrix arrangement after conversion, data logic correlation operation module carries out matrix data operation, data marking module carries out business data marking, and pushes it to data attribute target identification module for fusion data identification and screening after completion, and pushes it to fusion database; Data analysis logic control module extracts fusion data, and uses OCSVM abnormal data detection control module and Kmeans abnormal data detection control module for analysis respectively; The data output by OCSVM abnormal data detection control module is separated by density separation module through Gaussian distribution statistics and calculation of data distribution variance, and the data set is classified by data high-dimensional detection module, the abnormal data is dispersed, and the abnormal data is marked by abnormal data marking module, and pushed to multi-dimensional abnormal data marking module; The data output by Kmeans abnormal data detection control module is pushed to clustering operation module for clustering separation, the convex set intersection operation is completed by convex data set logic operation module, the set of n-dimensional positive semi-definite matrix is formed, the convex data set is converged by Kmeans convergence algorithm, the converged data set is separated, and pushed to multi-dimensional abnormal data marking module; The multi-dimensional abnormal data marking module performs intersection operation to form a multi-dimensional abnormal data set, and pushes the multi-dimensional abnormal data set to a multi-dimensional abnormal data logical judgment module to perform data format and protocol operation, and pushes the multi-dimensional abnormal data set to a service core control layer in the form of a message; The abnormal service research and judgment layer includes: an abnormal data triggering module that calls a Shark abnormal service shunting module to separate abnormal data logical attributes and push the abnormal data logical attributes to an abnormal service analysis operation module to perform service logic classification and definition, a thread operation module that performs service classification and flow transfer, and according to different service logic attributes, flows to a fault repair plan logical control module, an active repair logical control module, and a planned power outage logical control module, calls an abnormal service marking display module to complete marking of corresponding abnormal service attributes, and returns to the service core control layer.

2. The multi-layer network information convergence analysis and collaborative filtering troubleshooting system of claim 1, wherein, The data acquisition and transmission layer includes: a data acquisition layer and a data monitoring layer. The data acquisition layer acquires and stores distribution network equipment sensing data and power grid business resource data. The data monitoring layer performs quality analysis and data transmission monitoring on the data, and displays and marks the abnormality, wherein the quality analysis includes data attribute extraction, format conversion, dimension definition, and rule checking.

3. The multi-layer network information convergence analysis and collaborative filtering troubleshooting system of claim 2, wherein, The data acquisition layer includes: an integrated database, a distribution network equipment sensing data acquisition unit, and a power grid business resource data acquisition unit. The distribution network equipment sensing data acquisition unit includes: a front-end sensing module, a sensing data cache module, and a sensing data conversion processing module. The power grid business resource data acquisition unit includes: a front-end acquisition module, a resource data cache module, and a resource data conversion processing module.

4. The multi-layer network information convergence analysis and collaborative filtering troubleshooting system of claim 2 or 3, wherein, The data monitoring layer includes a quality monitoring unit and a transmission monitoring unit. The quality monitoring unit includes a data quality monitoring analysis module, a benchmark database, a data dimension definition statistics module, a data detection module, a checking module, and an analysis result display module. The data quality monitoring analysis module extracts data collected by the data acquisition layer and benchmark data in the benchmark database according to an instruction. The data detection module extracts data attributes. The data dimension definition statistics module performs format conversion and dimension definition on the data. The checking module performs rule checking on the data processed by the data dimension definition statistics module. The analysis result display module displays the rule checking result. The transmission monitoring unit comprises a token bucket data transmission monitoring module, a double-rate three-color marking module, an information rate calculation module, a message analysis and judgment module, and a token bucket marking cache module.

5. The multi-layer network information convergence analysis and collaborative filtering troubleshooting system of claim 1, wherein, The work order triggering layer, the work order triggering unit calls the Jbpm work order scheduling control module, pre-generates a business work order label according to the abnormal business maintenance scheme, the work order logic control unit calls the corresponding fault repair work order control module, the active repair work order control module, and the planned power outage work order control module according to the work order label, pushes the work order data and the work order label to the work order data marking unit for logical linkage, the work order data verification module verifies the work order data and the work order label with the abnormal data, pushes the work order data and the work order label to the work order data verification cache after verification is successful, the work order data verification logic judgment module performs logical verification operation on the business label and the work order label, generates a maintenance work order after completion, and returns to the service core control layer.

6. The multi-layer network information convergence analysis and collaborative filtering maintenance system of claim 1 or 2 or 3 or 5, characterized in that, The resource control layer comprises a Calibrated boosted trees cloud computing module, a logical operation module, a virtual resource management module, a user permission module, and a workflow management module, the Calibrated boosted trees cloud computing module calls the logical operation module to control the virtual resource management module, the user permission module, and the workflow management module respectively to perform engine resource management, user permission control, and workflow management respectively; Or / and, The big data analysis and display layer, the data preprocessing module pushes the marked abnormal business attributes and data to the data preprocessing module for preprocessing, pushes to the metadata management module and the data image conversion module after completion, performs data definition and data image conversion respectively, the data management module pushes the data to the FFT data operation module and the EM Training data operation module for data analysis respectively, pushes to the BI interactive module after completion, the data image conversion module pushes the image data to the data visualization analysis logical control unit for logical association of image data and data attributes, pushes to the BI interactive module after completion, the BI interactive module performs structured and unstructured data application classification and marking, pushes to the visualization display control module after completion, and performs data and big data graphics display through the perception layer data visualization model and the power grid business resource data visualization model; Or / and, The service core control layer, the service center total control unit completes data bidirectional transmission and instruction control through the FCFS scheduling unit, and stores data to the center database, and the FCFS scheduling unit calls corresponding logic control units of each layer to realize instruction sending, running control and data transmission between each layer and the service core control layer.

7. A method of use applied to the system of any one of claims 1 to 6, characterized in that, Comprise: The service core control layer controls the initialization of the resource control layer, the data acquisition transmission layer, the exception analysis processing layer and the big data analysis display layer; In response to the initialization success instruction, the service core control layer calls the resource control layer to perform engine resource management, user permission control and workflow management; In response to the resource control layer working normally, the service core control layer calls the data acquisition transmission layer to acquire distribution network equipment sensing data and power grid business resource data, and monitors data quality and transmission process; The service core control layer calls the exception analysis processing layer to identify and filter abnormal data in the distribution network equipment sensing data and the power grid business resource data by using various algorithms, and performs abnormal business research and judgment according to the abnormal data; The service core control layer calls the big data analysis display layer to generate an abnormal business maintenance scheme corresponding to the abnormal business; The service core control layer calls the exception analysis processing layer to generate a maintenance work order based on the abnormal business maintenance scheme; The service core control layer calls the big data analysis display layer to perform visual display of data.

8. An electronic device, comprising: Comprise a processor and a memory, the memory has a computer program stored therein, the computer program is loaded and executed by the processor to realize the method of claim 7. Comprise a processor and a memory, the memory has a computer program stored therein, the computer program is loaded and executed by the processor to realize the method of claim 7.

Images