A power distribution station defect identification system based on big data

Abstract

The application discloses a power distribution station defect identification system based on big data, and belongs to the technical field of power distribution station detection, and comprises a standard module, a model module, a collection route module, a control module, an analysis module and a server. The standard module is used for establishing a standard library. The model module is used for model management of the power distribution station, and obtains a power distribution station model. The collection route module is used for planning a collection route of a patrol unmanned aerial vehicle, and obtains a patrol route. The control module is used for controlling the patrol unmanned aerial vehicle to collect data, obtaining the patrol route of the patrol unmanned aerial vehicle, controlling the patrol unmanned aerial vehicle to fly according to the obtained patrol route, collecting data in a corresponding collection area when the patrol unmanned aerial vehicle reaches a corresponding stopping point, obtaining a corresponding collection image, marking a corresponding area table label, and integrating all the collected collection images into patrol data until the whole patrol route is flown. The analysis module is used for analyzing the patrol data, and judging whether defects exist.

Description

Technical Field

[0001] This invention belongs to the field of power distribution station detection technology, specifically a power distribution station defect identification system based on big data. Background Technology

[0002] With the rapid development of science and technology, people's demand for electricity is increasing, which promotes the rapid development of power plants and distribution stations. As the number of distribution stations increases, and some distribution stations are located in areas with fewer people, manual methods for identifying and inspecting distribution equipment and insulators in distribution stations are inefficient and inconvenient. Therefore, this invention provides a big data-based distribution station defect identification system that combines image acquisition by drones with intelligent identification and analysis to solve the problem of distribution station defect identification. Summary of the Invention

[0003] To address the problems of the above solutions, this invention provides a power distribution station defect identification system based on big data.

[0004] The objective of this invention can be achieved through the following technical solutions:

[0005] A big data-based substation defect identification system includes a standard module, a model module, a data acquisition route module, a control module, an analysis module, and a server.

[0006] The standard module is used to establish a standard library; the model module is used to manage the models of substations and obtain substation models.

[0007] The data collection route module is used to plan the data collection route of the inspection drone and obtain the inspection route;

[0008] The control module is used to control the inspection drone to collect data, obtain the inspection route of the inspection drone, control the inspection drone to fly according to the obtained inspection route, and when the inspection drone reaches the corresponding stop point, it collects data on the corresponding collection area, obtains the corresponding collection image, and marks the corresponding area table label, until the entire inspection route is flown, and integrates all the collected collection images into inspection data.

[0009] The analysis module is used to analyze the inspection data, obtain the inspection data, break down the inspection data into corresponding collected images, obtain the verification data corresponding to the collected images from the standard library, and analyze the collected images and verification data to determine whether there are defects.

[0010] Furthermore, the working method of the standard module includes:

[0011] Obtain standard test data for nameplates and insulators of various types of power distribution equipment in the substation. Segment the obtained standard test data according to the corresponding power distribution equipment to obtain verification data. Tag the verification data with the corresponding power distribution equipment and establish a first database. Input the verification data into the first database for storage. Mark the current first database as the standard library and update the verification data according to the updates of the corresponding specifications.

[0012] Furthermore, the working method of the model module includes:

[0013] Acquire a 3D data model of the substation and label it as the substation model. Mark the power distribution equipment to be inspected within the substation model and label it as the target equipment. Obtain the equipment information of the target equipment and match the corresponding verification data from the standard library based on the obtained equipment information. Based on the obtained verification data, mark the corresponding collection area on the corresponding power distribution equipment model in the substation model and mark the corresponding inspection type on the collection area to complete the model management of the substation.

[0014] Furthermore, the working method of the data acquisition route module includes:

[0015] The system acquires a substation model, identifies the location of each data collection area, sets corresponding stop points based on the identified data collection area locations, and marks the set stop points at their corresponding positions in the substation model. It also acquires the starting launch position of the inspection drone in real time, marks it as the starting point, marks the starting point accordingly within the substation model, and analyzes each stop point and starting point within the current substation model to obtain the corresponding inspection route.

[0016] Furthermore, the method for setting corresponding stop points based on the identified collection area location includes:

[0017] Obtain the collection range of the inspection drone, mark the obtained collection range as the reference range, compare the reference range with each collection area, and divide the collection area into the first area and the second area according to the comparison results. Analyze the first area and set a stop point, which is marked as the first stop point. Analyze the second area and set several stop points, which are marked as the second stop points.

[0018] Furthermore, the method for analyzing and setting a stop point for the first area includes:

[0019] The shape of the corresponding collection area is marked as the first shape. The reference area is covered on the first shape, and the diameter of the reference area is gradually reduced. When the boundary of the reference area touches the first shape, the center position of the reference area is adjusted, and the diameter of the reference area is reduced again until the reference area can no longer be reduced. The center of the current reference area is identified, and the corresponding dwell point is calculated based on the identified center of the reference area.

[0020] Furthermore, the methods for analyzing each dwell point and starting point within the current substation model include:

[0021] Identify several second stop points corresponding to a second region, set a representative point based on the second stop points, and mark the representative point in the substation model accordingly. Generate several candidate routes based on the starting point, representative point, and first stop point, and mark each route segment within the candidate routes as i, i = 1, 2, ..., n, where n is a positive integer; identify the length of each route segment and mark it as LXi; identify the direction of travel of the corresponding candidate route through the representative point, set the segment correction value and mark it as XZi, and obtain the corresponding internal routes according to the formula. Calculate the priority value, select the corresponding candidate route as the target route based on the calculated priority value, and combine the target route with the internal routes to obtain the inspection route.

[0022] Furthermore, the method for selecting the corresponding candidate route as the target route based on the calculated priority value is as follows:

[0023] The calculated priority values ​​are sorted in ascending order to obtain the first sequence. The candidate routes corresponding to the first sequence are identified and marked as the target routes.

[0024] Furthermore, methods for combining the target route with internal routes include:

[0025] Identify representative points in the target route, match them with the corresponding internal routes, replace the representative points with the internal routes, identify the entry and exit points of the internal routes, connect the corresponding points, and obtain the inspection route.

[0026] Compared with the prior art, the beneficial effects of the present invention are: through the cooperation between the standard module, model module, data acquisition route module, control module and analysis module, intelligent substation defect detection is achieved, which greatly improves detection efficiency, reduces the inspection burden of corresponding personnel, generates corresponding inspection routes according to the actual situation of the substation, and intelligently controls the inspection drone. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a block diagram illustrating the principle of the present invention. Detailed Implementation

[0029] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0030] like Figure 1 As shown, a big data-based substation defect identification system includes a standard module, a model module, a data acquisition route module, a control module, an analysis module, and a server.

[0031] The standard module is used to establish a standard library, and the specific methods include:

[0032] Based on big data analysis, standard test data for various types of power distribution equipment labels and insulators within the substation are obtained, such as installation location ranges, data on specified markings, and standard data on insulator appearance. The obtained standard test data is then broken down according to the corresponding power distribution equipment to obtain verification data. For example, verification data for labels refers to the standard requirements that a label for a certain power distribution equipment should meet. The verification data is then tagged with the corresponding power distribution equipment, and a first database is established. The verification data is then input into the first database for storage, and the current first database is marked as a standard library. The verification data is updated accordingly based on updates to the relevant standards.

[0033] The model module is used for model management of substations, and the specific methods include:

[0034] Acquire a 3D data model of the substation and mark it as the substation model. For substation 3D data models that cannot be directly obtained, a new 3D data model of the substation can be rebuilt based on existing modeling techniques. Mark the power distribution equipment that needs to be inspected within the substation model and mark it as the target equipment. Obtain the equipment information of the target equipment, such as type and model. Match the corresponding verification data from the standard library based on the obtained equipment information. Mark the corresponding collection area on the corresponding power distribution equipment model in the substation model based on the obtained verification data, and mark the corresponding inspection type on the collection area. The inspection type refers to whether it is an inspection label or an inspection insulator, thus completing the model management of the substation.

[0035] Based on the obtained verification data, the corresponding data collection area is marked on the power distribution equipment model in the power distribution station model. This means that according to the standard installation requirements of the corresponding signs or insulators in the verification data, the corresponding installation interval is determined and the corresponding location interval is marked, which is the data collection area. When collecting data in the future, the data can be collected directly in the data collection area, which can greatly reduce the amount of data to be collected and the amount of data analysis.

[0036] The data collection route module is used to plan the data collection route of the inspection drone, and the specific methods include:

[0037] The system acquires a substation model, identifies the locations of various data collection areas, and sets corresponding stopping points based on these locations. A stopping point refers to the point where the inspection drone stops when collecting information within the corresponding data collection area. These stopping points are then marked at their corresponding positions in the substation model. The system also acquires the drone's initial takeoff position in real time and marks it as the starting point. This starting point is then marked accordingly within the substation model. Finally, the system analyzes each stopping point and starting point within the current substation model to obtain the corresponding inspection route.

[0038] Methods for setting corresponding stop points based on the identified collection area location include:

[0039] Obtaining the collection range of the inspection drone refers to the maximum effective collection range of the inspection drone while meeting requirements such as collection clarity. The obtained collection range is marked as the reference range. The reference range is compared with each collection area. Based on the comparison results, the collection area is divided into the first area and the second area. The first area refers to the collection area where the reference area is not smaller than the collection area, meaning that one stop point can completely collect information within the entire collection area. The second area refers to the collection area where the reference area is smaller than the collection area, requiring multiple stop points to be set for collection. Analyze the first area and set one stop point, marking it as the first stop point. Analyze the second area and set several stop points, marking them as the second stop points.

[0040] Methods for setting a stop point for analyzing the first region include:

[0041] The shape of the corresponding collection area is marked as the first shape. The reference area is covered on the first shape, and the diameter of the reference area is gradually reduced. When the boundary of the reference area touches the first shape, the center position of the reference area is adjusted, and the diameter of the reference area is reduced again until the reference area can no longer be reduced. This means that when reducing the size further, no matter how the position of the reference area is adjusted, the boundary of the first shape will exceed the reference area. The center of the current reference area is identified, and the corresponding dwell point is calculated based on the identified center of the reference area. The corresponding straight-line distance can be determined according to the current calculation method, and then the corresponding dwell point can be determined.

[0042] The method for analyzing and setting several stop points in the second area is as follows:

[0043] A point analysis model is established based on a CNN or DNN network. The model is trained by manually setting up the corresponding training set to obtain the corresponding collection area information and the reference area information of the inspection drone. The successfully trained point analysis model is then used for analysis to obtain the corresponding stopping points. The specific establishment and training process is common knowledge in this field, so it will not be described in detail.

[0044] Methods for analyzing various dwell points and starting points within the current substation model include:

[0045] Identify several second stop points corresponding to a second region, set a representative point based on the second stop points, and mark the representative point in the substation model accordingly. Generate several candidate routes based on the starting point, representative point, and first stop point, and mark each route segment within the candidate routes as i, i = 1, 2, ..., n, where n is a positive integer; identify the length of each route segment and mark it as LXi; identify the direction of travel of the corresponding candidate route through the representative point, set the segment correction value and mark it as XZi, and obtain the corresponding internal routes according to the formula. Calculate the priority value, select the corresponding candidate route as the target route based on the calculated priority value, and combine the target route with the internal routes to obtain the inspection route.

[0046] Set a representative point based on the second stop point. That is, set a representative point based on the locations of several second stop points in the second area. You can set a center point or select a second stop point as the representative point. You can make corresponding adjustments according to actual needs.

[0047] The method for generating several candidate routes based on the starting point, representative point, and first stop point is as follows: a route model is built based on a CNN or DNN network, and the corresponding training set is set manually for training. The successfully trained route model is analyzed to generate possible candidate routes.

[0048] The method for identifying the direction of travel of the candidate route passing through representative points and setting road segment correction values ​​includes: determining the nearest second stop point based on the corresponding direction of travel and marking it as the entry point; determining the exit point based on the entry and exit points; planning the routes of each current second stop point based on the entry and exit points and marking them as internal routes, and planning according to the shortest distance. The determination of the entry and exit points can be achieved using existing technology, so it will not be described in detail. A correction model is built based on a CNN or DNN network, and a corresponding training set is manually set for training. Based on changes in the corresponding road segments, corresponding road segment correction values ​​are set, and a training set is built. The successfully trained correction model analyzes route segments with representative points and outputs the corresponding road segment correction values. For route segments without representative points, no road segment correction value is output.

[0049] The method for selecting the corresponding candidate route as the target route based on the calculated priority value is as follows:

[0050] The calculated priority values ​​are sorted in ascending order to obtain the first sequence. The candidate routes corresponding to the first sequence are identified and marked as the target routes.

[0051] Methods that combine the target route with internal routes include:

[0052] Identify representative points in the target route, match them with the corresponding internal routes, replace the representative points with the internal routes, identify the entry and exit points of the internal routes, connect the corresponding points, and obtain the inspection route.

[0053] The control module is used to control the inspection drone to collect data, and the specific methods include:

[0054] The inspection route of the inspection drone is obtained, and the drone is controlled to fly according to the obtained route. When the drone reaches the corresponding stop point, it collects data in the corresponding collection area, obtains the corresponding collection image, and marks the corresponding area table. This process continues until the entire inspection route is completed, and all the collected images are integrated into inspection data.

[0055] The analysis module is used to analyze the inspection data, and the specific methods include:

[0056] The inspection data is obtained and broken down into corresponding images. Verification data corresponding to the images is retrieved from the standard library. The images and verification data are analyzed to determine whether there are defects.

[0057] Based on the collected images and verification data, a verification analysis model is established using a CNN or DNN network. The model is trained manually using a corresponding training set. Combined with image recognition technology, it is used to determine whether there are defects such as missing signs, incorrect markings, or cracked insulators. The successfully trained verification analysis model is then used to analyze the collected images and verification data to obtain the corresponding analysis results.

[0058] The above formulas are all numerical calculations after removing dimensions. The formulas are obtained by software simulation based on a large amount of data and are closest to the real situation. The preset parameters and preset thresholds in the formulas are set by those skilled in the art according to the actual situation or obtained by simulation based on a large amount of data.

[0059] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A big data-based substation defect identification system, characterized in that, It includes a standard module, a model module, a data acquisition route module, a control module, an analysis module, and a server; The standard module is used to establish a standard library; the model module is used to manage the models of substations and obtain substation models. The data collection route module is used to plan the data collection route of the inspection drone and obtain the inspection route; The control module is used to control the inspection drone to collect data, obtain the inspection route of the inspection drone, control the inspection drone to fly according to the obtained inspection route, and when the inspection drone reaches the corresponding stop point, it collects data on the corresponding collection area, obtains the corresponding collection image, and marks the corresponding area table label, until the entire inspection route is flown, and integrates all the collected collection images into inspection data. The analysis module is used to analyze the inspection data, obtain the inspection data, break down the inspection data into corresponding collected images, obtain the verification data corresponding to the collected images from the standard library, and analyze the collected images and verification data to determine whether there are defects. The working method of the data acquisition route module includes: The system acquires a substation model, identifies the location of each data collection area, sets corresponding stopping points based on the identified data collection area locations, and marks the set stopping points at their corresponding positions in the substation model. It also acquires the starting launch position of the inspection drone in real time, marks it as the starting point, marks the starting point in the substation model, and analyzes each stopping point and starting point in the current substation model to obtain the corresponding inspection route. Methods for setting corresponding stop points based on the identified collection area location include: Obtain the collection range of the inspection drone, mark the obtained collection range as the reference range, compare the reference range with each collection area, divide the collection area into the first area and the second area according to the comparison results, analyze the first area and set a stop point, mark it as the first stop point, analyze the second area and set several stop points, mark them as the second stop points; Methods for analyzing various dwell points and starting points within the current substation model include: Identify several second stop points corresponding to a second region, set a representative point based on the second stop points, and mark the representative point accordingly in the substation model. Generate several candidate routes based on the starting point, representative point, and first stop point, and mark each route segment within the candidate routes as i, i=1, 2, ..., n, where n is a positive integer; identify the length of each route segment and mark it as LXi; identify the direction of travel of the corresponding candidate route through the representative point, set the segment correction value and mark it as XZi, and obtain the corresponding internal route according to the formula. Calculate the priority value, select the corresponding candidate route as the target route based on the calculated priority value, and combine the target route with the internal routes to obtain the inspection route.

2. The big data-based substation defect identification system according to claim 1, characterized in that, The standard module's working methods include: Obtain standard test data for nameplates and insulators of various types of power distribution equipment in the substation. Segment the obtained standard test data according to the corresponding power distribution equipment to obtain verification data. Tag the verification data with the corresponding power distribution equipment and establish a first database. Input the verification data into the first database for storage. Mark the current first database as the standard library and update the verification data according to the updates of the corresponding specifications.

3. The big data-based substation defect identification system according to claim 1, characterized in that, The working methods of the model module include: Acquire a 3D data model of the substation and label it as the substation model. Mark the power distribution equipment to be inspected within the substation model and label it as the target equipment. Obtain the equipment information of the target equipment and match the corresponding verification data from the standard library based on the obtained equipment information. Based on the obtained verification data, mark the corresponding collection area on the corresponding power distribution equipment model in the substation model and mark the corresponding inspection type on the collection area to complete the model management of the substation.

4. The big data-based substation defect identification system according to claim 1, characterized in that, Methods for setting a stop point for analyzing the first region include: The shape of the corresponding collection area is marked as the first shape. The reference area is covered on the first shape, and the diameter of the reference area is gradually reduced. When the boundary of the reference area touches the first shape, the center position of the reference area is adjusted, and the diameter of the reference area is reduced again until the reference area can no longer be reduced. The center of the current reference area is identified, and the corresponding dwell point is calculated based on the identified center of the reference area.

5. The big data-based substation defect identification system according to claim 1, characterized in that, The method for selecting the corresponding candidate route as the target route based on the calculated priority value is as follows: The calculated priority values ​​are sorted in ascending order to obtain the first sequence. The candidate routes corresponding to the first sequence are identified and marked as the target routes.

6. The big data-based substation defect identification system according to claim 5, characterized in that, Methods that combine the target route with internal routes include: Identify representative points in the target route, match them with the corresponding internal routes, replace the representative points with the internal routes, identify the entry and exit points of the internal routes, connect the corresponding points, and obtain the inspection route.

Images