An intelligent unmanned system and a flying inspection method for power grid distribution room inspection

Abstract

The application discloses an intelligent unmanned system and a flight inspection method for power grid distribution room inspection, relates to the technical field of distribution room inspection, and solves the technical problem that the distribution room inspection is completed by manual inspection, which consumes a large amount of human resources. By applying intelligent equipment and algorithms, the traditional manual operation mode is replaced, the occupation of human resources is greatly reduced, the error rate under frequent repeated operation of personnel is avoided, the inspection frequency is improved, the fault risk is reduced, the shooting data is processed and analyzed in real time, different levels of alarms are completed through analysis results, so that external personnel can timely understand the situation inside the distribution room space, timely action corresponding protection measures, the inspection data is classified and archived, the fault and data traceability and data sharing are facilitated, the inspection results are opened, and are provided for multiple use departments to avoid repeated operation and reduce operation cost.

Description

Technical Field

[0001] This invention belongs to the field of power distribution room inspection technology, specifically relating to an intelligent unmanned system and a flight inspection method for power grid power distribution room inspection. Background Technology

[0002] With the continuous integration of smart grids and smart manufacturing, the power distribution room, as a core facility of the power grid, will inevitably bear the task of regional core nodes and shoulder the important responsibility of ensuring power supply and smooth operation. Therefore, the daily inspection of the power distribution room is particularly important. However, this work is a simple repetitive task with no technical difficulty, only tedious checking and recording. At present, this work is basically completed by manual inspection, which consumes a lot of human resources.

[0003] To explore fully automated autonomous indoor inspection technology to completely replace manual inspections, it is proposed to integrate multiple intelligent technologies for comprehensive inspection of indoor power distribution rooms. Through data relay equipment, the inspection results and meter reading data will be uploaded to the backend in real time using communication technology to meet daily routine inspection and meter reading work. It can analyze and store inspection data in real time, realize real-time monitoring, early warning, analysis and visualization of the station, and provide support for the refined management of the power grid. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent unmanned system and a flight inspection method for power grid distribution room inspection, so as to solve the problem that the daily inspection of distribution rooms in the prior art is completed by manual inspection, which consumes a lot of human resources.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides an intelligent unmanned system for inspecting power grid distribution rooms, comprising several sets of fixed base stations, an intelligent on-site inspection system for distribution rooms, and a remote intelligent identification system;

[0007] The on-site intelligent inspection system includes an inspection drone, an indoor high-precision positioning module, and an on-site command and dispatch station; the remote intelligent identification system includes a computing server and an intelligent identification terminal; and the several sets of fixed base stations are evenly distributed within the power distribution room.

[0008] The on-site command and dispatch station is used to receive temporary inspection instructions and fixed inspection time settings from the remote terminal. After receiving the temporary inspection instructions from the on-site command and dispatch station or reaching the fixed inspection time point, the inspection drone takes off autonomously on the take-off and landing platform. The on-site command and dispatch station is located in the take-off and landing platform. The inspection drone uses an indoor high-precision positioning module and camera unit for positioning and shooting. After real-time calculation by the onboard computing unit, it obtains its own high-precision position information and performs autonomous indoor flight based on the high-precision position information. According to the preset flight inspection process, the inspection drone hovers and flies at each inspection point.

[0009] As an optional technical solution of the present invention, the flight inspection process planning steps of the inspection drone are as follows:

[0010] S1. Label the different distribution cabinets sequentially, using the label format B. i Where i represents different power distribution cabinets, and i = 1, 2, ..., n;

[0011] S2. Arrange the markers Bi according to B1, B2, ..., B n The distribution cabinets are arranged sequentially in the following manner. During the sorting process, when a diagonal turn is encountered, mark B. j And B j-1 Add marker Z to make B j The form of expression is: B jz Let j and j-1 both belong to i. The marker for a left turn is Z; the marker for a right turn is Y. If turning right, B... j The form of expression is B jY ;

[0012] S3. Arrange the corresponding mark forms in the recognition order in the onboard computing unit inside the inspection drone, and at the same time limit the inspection speed, inspection altitude and shooting distance of the inspection drone. When the mark with Z and Y is identified, set the corresponding turning command to make the inspection drone turn left or right.

[0013] S4. When the camera unit inside the inspection drone is fully aligned with the panel image inside the instrument, it directly captures and extracts the image, and transmits the captured high-definition image to the on-site command and dispatch station in real time.

[0014] As an optional technical solution of the present invention, the on-site command and dispatch station includes a data industrial control computer and a 5G routing base station. The data industrial control computer is used to store the captured high-definition images, and the 5G routing base station is used to send the high-definition images to the remote intelligent identification system.

[0015] As an optional technical solution of the present invention, the remote intelligent identification system is used to intelligently identify high-definition images, write the identification results into a report, and generate a report on the results of the inspection after the drone has completed the inspection of the entire power distribution room. Simultaneously, the raw data and the report are classified and stored, with each power distribution cabinet matched with a corresponding storage space, and different storage spaces using corresponding labels. i Mark the abnormalities found during the inspection process and trigger an alarm based on the severity level of the abnormality.

[0016] As an optional technical solution of the present invention, the steps of the remote intelligent identification system to perform an alarm operation according to the severity level of the anomaly are as follows:

[0017] W1, the intelligent recognition terminal acquires high-definition images and labels them as GQ. i 'i' represents different power distribution cabinets;

[0018] W2, for high-definition images GQ i Internal rotation speed, temperature, and power consumption; temperature information is pre-acquired and tagged as W. i The transport server has a corresponding temperature threshold Q set inside, when W i When W ≤ Q, the computation server proceeds to the next calculation step; when W ≤ Q, the computation server proceeds to the next calculation step. i When the value is greater than Q, the computing server generates a level 1 alarm signal and sends it to an external terminal for display, and controls the external alarm terminal to emit a level 1 alarm sound.

[0019] W3, Set the speed to Z i Set the power consumption to GH. i ,use Obtain alarm value BJ i C1 and C2 are both preset fixed coefficient factors;

[0020] W4, where the transport server includes a secondary alarm threshold R, when BJ i When the value is greater than or equal to R, a secondary alarm signal is generated and sent to an external terminal for display. At the same time, the alarm terminal is controlled to emit a secondary alarm ring.

[0021] As an optional technical solution of the present invention, the computing server includes a feature extraction module and a feature pyramid fusion module. The feature extraction module first extracts feature images from several high-definition images and transmits the extracted feature images to the feature pyramid fusion module. The feature pyramid fusion module samples the feature images layer by layer and fuses them with the feature images of the previous layer to generate a report.

[0022] As an optional technical solution of the present invention, the first-level alarm ringtone and the second-level alarm ringtone correspond to different action plans.

[0023] As an optional technical solution of the present invention, the inspection drone is equipped with an anti-collision outer protective cover.

[0024] As an optional technical solution of the present invention, the fixed base station is wall-mounted and fixedly installed in the power distribution room.

[0025] In a second aspect, the present invention provides a flight inspection method for the aforementioned intelligent unmanned system, comprising the following steps:

[0026] S1. Label the different distribution cabinets sequentially, using the label format B. i Where i represents different power distribution cabinets, and i = 1, 2, ..., n;

[0027] S2. Arrange the markers Bi according to B1, B2, ..., B n The distribution cabinets are arranged sequentially in the following manner. During the sorting process, when a diagonal turn is encountered, mark B. j And B j-1 Add marker Z to make B j The form of expression is: B jz Let j and j-1 both belong to i. The marker for a left turn is Z; the marker for a right turn is Y. If turning right, B... j The form of expression is B jY ;

[0028] S3. Arrange the corresponding mark forms in the recognition order in the onboard computing unit inside the inspection drone, and at the same time limit the inspection speed, inspection altitude and shooting distance of the inspection drone. When the mark with Z and Y is identified, set the corresponding turning command to make the inspection drone turn left or right.

[0029] S4. When the camera unit inside the inspection drone is fully aligned with the panel image inside the instrument, it directly captures and extracts the image, and transmits the captured high-definition image to the on-site command and dispatch station in real time.

[0030] Compared with the prior art, the beneficial effects of the present invention are:

[0031] 1) By applying intelligent equipment and algorithms, the traditional manual operation method is replaced, which greatly reduces the occupation of human resources, avoids the increase in error rate due to frequent repetitive work, increases the frequency of inspections, reduces the risk of failure, and processes and analyzes the captured data in real time. The analysis results are used to complete alarms of different levels, so that external personnel can understand the situation inside the power distribution room in a timely manner and take corresponding protective measures.

[0032] 2) Categorizing and archiving inspection data facilitates fault and data tracing and data sharing. Opening up inspection results allows multiple user departments to avoid duplication of work and reduce operating costs. Attached Figure Description

[0033] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0034] Figure 1 This is a schematic diagram of the principle framework of the present invention;

[0035] Figure 2 This is a schematic diagram illustrating the analysis and recognition of high-definition images according to the present invention;

[0036] Figure 3 This is a schematic diagram of the operation process of the inspection drone of the present invention. Detailed Implementation

[0037] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0038] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used in this invention is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0039] Please see Figure 1 This application provides an intelligent unmanned system for power grid distribution room inspection, including a distribution room on-site intelligent inspection system and a remote intelligent identification system;

[0040] The on-site intelligent inspection system consists of an inspection drone, an indoor high-precision positioning module, and an on-site command and dispatch station; the remote intelligent identification system consists of a computing server and an intelligent identification terminal.

[0041] Furthermore, the intelligent inspection system for the power distribution room also includes several sets of fixed base stations, which are evenly distributed within the interior space of the power distribution room. The fixed base stations are wall-mounted and do not occupy ground space. Each base station is connected to 220V mains power and can work continuously after calibration.

[0042] The on-site command and dispatch station receives temporary inspection instructions and fixed inspection time settings from the remote terminal (the temporary inspection instructions and fixed inspection time are determined by external operators). After receiving the temporary inspection instructions from the on-site command and dispatch station or reaching the fixed inspection time point, the inspection drone takes off autonomously on the take-off and landing platform. The on-site command and dispatch station is located in the take-off and landing platform. The inspection drone uses an indoor high-precision positioning module and camera unit for positioning and shooting. The onboard computing unit calculates and obtains its own high-precision position information in real time. It achieves the ability to fly autonomously indoors through high-precision position information. According to the on-site physical environment, a flight inspection process is planned. The inspection drone hovers and flies at each inspection point.

[0043] The flight inspection process planning steps are as follows:

[0044] S1. Label the different distribution cabinets sequentially, using the label format B. i Where i represents different power distribution cabinets, and i = 1, 2, ..., n;

[0045] S2. Arrange the markers Bi according to B1, B2, ..., B n The distribution cabinets are arranged sequentially in the following manner. During the sorting process, when a diagonal turn is encountered, mark B. j And B j-1 Add marker Z to make B j The form of expression is: B jz Let , where j and j-1 both belong to i, the left turn is marked as Z, the right turn is marked as Y, and if turning right, then B j The form of expression is: B jY (where each different label B) i (These are all different inspection points);

[0046] S3. The operator arranges the corresponding markings in the identification order and inspects the onboard computing unit inside the inspection drone. At the same time, the inspection speed, inspection altitude and shooting distance of the inspection drone are limited. When the markings with Z and Y are identified, the corresponding turning command is set to make the inspection drone turn left or right.

[0047] S4. When the camera unit inside the inspection drone is fully aligned with the panel image inside the instrument, it directly captures and extracts the image, and transmits the captured image to the on-site command and dispatch station in real time.

[0048] The inspection drone includes a mobile beacon unit, a data transmission unit, a camera unit, and an onboard computing unit. The mobile beacon unit is used to obtain high-precision positioning information from several fixed base stations. The camera unit takes pictures of the corresponding instrument panel to obtain high-definition images of the instrument panel, and then sends the high-definition images to the industrial control computer of the command and dispatch station for storage in real time through the data transmission unit.

[0049] The on-site command and dispatch station includes data transmission unit two, a data industrial control computer, and a 5G routing base station;

[0050] The second data transmission unit is used to acquire flight commands and high-definition images sent by the first data transmission unit. The data control computer of the on-site command and dispatch station stores the high-definition images and sends the image files to the remote intelligent identification system through the 5G routing base station deployed on-site.

[0051] After receiving high-definition images, the remote intelligent recognition system performs intelligent recognition on the high-definition images and writes the recognition results into a report. After the drone finishes inspecting the entire power distribution room, it generates a report on the results of this inspection. At the same time, the raw data and reports are classified and stored for subsequent manual review. For any abnormalities found during the inspection, the system will trigger an alarm operation according to the severity level of the abnormality. After completing the inspection task, the drone will autonomously return to the take-off and landing platform to recharge and prepare for the next inspection task.

[0052] The steps for the remote intelligent identification system to trigger an alarm based on the severity level of the anomaly are as follows:

[0053] W1, the intelligent recognition terminal acquires high-definition images and labels them as GQ. i , i represents different power distribution cabinets, where i = 1, 2, ..., n;

[0054] W2, for high-definition images GQ i Internal rotation speed, temperature, and power consumption; temperature information is pre-acquired and tagged as W. i The transport server has a corresponding temperature threshold Q set inside, when W i When W ≤ Q, the computation server proceeds to the next calculation step; when W ≤ Q, the computation server proceeds to the next calculation step. i When the value is greater than Q, the computing server generates a level 1 alarm signal and sends it to an external terminal for display, and controls the external alarm terminal to emit a level 1 alarm sound.

[0055] W3, Set the speed to Z i Set the power consumption to GH. i ,use Obtain alarm value BJ i C1 and C2 are both preset fixed coefficient factors;

[0056] W4, where the transport server includes a secondary alarm threshold R, when BJ i When the value is greater than or equal to R, a secondary alarm signal is generated and sent to an external terminal for display. At the same time, the alarm terminal is controlled to emit a secondary alarm ring.

[0057] There are different action plans for the Level 1 and Level 2 alarms. Specifically, the Level 1 alarm indicates a greater level of danger than the Level 2 alarm. When external maintenance personnel hear the Level 1 alarm, they should immediately evacuate the area, wear appropriate clothing, and proceed with inspection and repair work. When they hear the Level 2 alarm, they should immediately repair the designated power distribution cabinet.

[0058] Inspection drones: Key performance indicators for drones were designed to adapt to the inspection environment inside power distribution rooms.

[0059]

[0060] Remote Intelligent Identification System: The remote intelligent identification system consists of a high-performance computing server and an intelligent identification algorithm running inside the server. The computing server is connected to the Internet and has a fixed public IP address. It receives basic information and high-definition images of the inspection drone sent back by the on-site command and dispatch station. After classifying and storing the high-definition images, the intelligent identification algorithm identifies and analyzes the images, writes the analysis results into data reports, generates reports and alarm information, and finally classifies and stores the reports. At the same time, the reports and alarm information can be sent to a designated address or client according to a pre-designed process.

[0061] Please see Figure 2 The steps for image recognition and analysis using intelligent recognition algorithms are as follows: the computing server includes a feature extraction module and a feature pyramid fusion module; the feature extraction module extracts feature images from several high-definition images in advance, and the feature pyramid fusion module upsamples the feature images layer by layer and fuses them with the feature images of the previous layer; the feature extraction module is DenseNet, and the feature pyramid fusion module includes four optimization units and one stitching unit, as well as 4x upsampling and classification.

[0062] It solves the multi-angle problem in object detection, can efficiently extract features from various dimensions of an image, and improves the accuracy of small object detection.

[0063] Please see Figure 3The inspection drone's inspection route is a counter-clockwise loop around the field, with a flight speed of approximately 2 meters per second, a flight altitude of 1.2 meters above the ground, and a distance of 1 meter between the route and the target. All task points are pre-set along the route, and the drone hovers at each task point to collect data. To ensure flight safety, the flight altitude is strictly maintained within ±0.2 meters of the preset altitude, and a no-fly zone is set within 0.5 meters of the target object, prohibiting the inspection drone from flying into this area. After completing the inspection task, the inspection drone returns to the take-off and landing platform along the preset route, automatically recharges after coming to a stop, and awaits the next operation.

[0064] The inspection drone take-off and landing platform (with an integrated on-site command and dispatch station) is located near the safety exit, in an open and unobstructed area, facilitating drone take-off and landing and personnel operation and maintenance. The platform has a flat top and is equipped with auxiliary landing markers and an automatic charging device. It is powered by 220V AC mains electricity. The on-site command and dispatch station can communicate with the drones in real time and forward data to a remote intelligent identification system via the network. Operators can also set and adjust drone operation data on-site through the on-site command and dispatch station's function setting panel. The platform also includes a material storage area for major consumables, vulnerable parts, and maintenance tools.

[0065] UWB technology is adopted as the positioning technology solution for the autonomous flight of inspection drones. The positioning system consists of base stations and beacons, with the beacons deployed on the inspection drones.

[0066] In the above-disclosed content, some of the data in the formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the real situation, obtained by software simulation of a large amount of collected data. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.

[0067] The working principle of this invention: The on-site command and dispatch station receives temporary inspection instructions and fixed inspection time settings from the remote terminal. Upon receiving a temporary inspection task message from the on-site command and dispatch station or reaching the fixed inspection time point, the UAV autonomously takes off from the take-off and landing platform. The on-site command and dispatch station is located within the take-off and landing platform. The UAV, through an indoor high-precision positioning module and an onboard vision camera, obtains its own high-precision position information in real time via an onboard computer. This high-precision position information enables autonomous flight indoors. Based on the on-site physical environment, a flight inspection process is planned. The UAV hovers at each inspection point, capturing high-definition images of the corresponding instrument panels using an onboard high-definition camera. This high-definition image is then transmitted in conjunction with the on-site command and dispatch station. The data transmission link between stations sends image files to the industrial control computer of the command and dispatch station in real time for storage. While storing the image files, the industrial control computer of the command and dispatch station sends the image files to the remote intelligent recognition system server through the 5G routing base station deployed on site. After receiving the image files, the recognition system performs intelligent recognition on the image files and writes the recognition results into a report. After the drone completes the inspection of the entire power distribution room, it generates a report on the results of this inspection. At the same time, the raw data and the report are classified and stored for subsequent manual review. For any abnormalities found during the inspection, the system will trigger an alarm operation according to the severity level of the abnormality. After completing the inspection mission, the drone will autonomously return to the take-off and landing platform to recharge and prepare for the next inspection mission.

[0068] In other embodiments, the present invention also provides a separate flight inspection method for the aforementioned intelligent unmanned system, comprising the following steps:

[0069] S1. Label the different distribution cabinets sequentially, using the label format B. i Where i represents different power distribution cabinets, and i = 1, 2, ..., n;

[0070] S2. Arrange the markers Bi according to B1, B2, ..., B n The distribution cabinets are arranged sequentially in the following manner. During the sorting process, when a diagonal turn is encountered, mark B. j And B j-1 Add marker Z to make B j The form of expression is: B jz Let j and j-1 both belong to i. The marker for a left turn is Z; the marker for a right turn is Y. If turning right, B... j The form of expression is B jY ;

[0071] S3. Arrange the corresponding mark forms in the recognition order in the onboard computing unit inside the inspection drone, and at the same time limit the inspection speed, inspection altitude and shooting distance of the inspection drone. When the mark with Z and Y is identified, set the corresponding turning command to make the inspection drone turn left or right.

[0072] S4. When the camera unit inside the inspection drone is fully aligned with the panel image inside the instrument, it directly captures and extracts the image, and transmits the captured high-definition image to the on-site command and dispatch station in real time.

[0073] As is known from common technical knowledge, this invention can be implemented through other embodiments that do not depart from its spirit or essential characteristics. Therefore, the disclosed embodiments described above are merely illustrative in all respects and are not the only ones. All modifications within the scope of this invention or its equivalents are included in this invention.

Claims

1. A method for aerial inspection of power grid distribution rooms, based on an intelligent unmanned system, characterized in that, Intelligent unmanned systems include on-site intelligent inspection systems for power distribution rooms and remote intelligent identification systems; The power distribution room intelligent inspection system includes an inspection drone, an indoor high-precision positioning module, and an on-site command and dispatch station; the remote intelligent identification system includes a computing server and an intelligent identification terminal. indoor The high-precision positioning module consists of several fixed base stations evenly distributed within the power distribution room. The on-site command and dispatch station is used to receive temporary inspection instructions and fixed inspection time settings from the remote intelligent identification system. After receiving the temporary inspection instructions from the on-site command and dispatch station or reaching the fixed inspection time point, the inspection drone takes off autonomously on the take-off and landing platform. The on-site command and dispatch station is located in the take-off and landing platform. The inspection drone uses an indoor high-precision positioning module and camera unit for positioning and shooting. After real-time calculation by the onboard computing unit, it obtains its own high-precision position information and performs autonomous indoor flight based on the high-precision position information. According to the preset flight inspection process, the inspection drone hovers and flies at each inspection point. The indoor high-precision positioning module uses UWB technology. The fixed base station is a UWB base station, which is wall-mounted on the wall of the power distribution room at a height higher than the maximum flight altitude of the inspection drone. The inspection drone includes a mobile beacon unit, which is used to obtain high-precision positioning information from several groups of fixed base stations. The flight inspection method includes the following steps: S1. Label the different distribution cabinets sequentially, using the label format B. i Where i represents different power distribution cabinets, and i = 1, 2, ..., n; S2. Arrange the markers Bi according to B1, B2, ..., B n The distribution cabinets are arranged sequentially in the following manner. During the sorting process, when a diagonal turn is encountered, mark B. j And B j-1 Add marker Z to make B j The form of expression is: B jz Let j and j-1 both belong to i. The marker for a left turn is Z; the marker for a right turn is Y. If turning right, B... j The form of expression is B jY ; S3. Arrange the corresponding mark forms into a recognition order and input them into the onboard computing unit inside the inspection drone. At the same time, limit the inspection speed, inspection altitude and shooting distance of the inspection drone. When the mark with Z and Y is recognized, set the corresponding turning command to make the inspection drone turn left or right. S4. When the camera unit inside the inspection drone is fully aligned with the panel image inside the instrument, it directly captures and extracts the image, and transmits the captured high-definition image to the on-site command and dispatch station in real time.

2. The flight inspection method according to claim 1, characterized in that, The on-site command and dispatch station includes a data industrial control computer and a 5G routing base station. The data industrial control computer is used to store the captured high-definition images, and the 5G routing base station is used to send the high-definition images to the remote intelligent recognition system.

3. The flight inspection method according to claim 2, characterized in that, The remote intelligent recognition system is used to intelligently recognize high-definition images, write the recognition results into a report, and generate a report on the inspection results after the drone has completed its inspection of the entire power distribution room. Simultaneously, the raw data and report are categorized and stored, with each power distribution cabinet matched with a corresponding storage space, and different storage spaces using a corresponding label B. i Mark the abnormalities found during the inspection process and trigger an alarm based on the severity level of the abnormality.

4. The flight inspection method according to claim 3, characterized in that, The computing server includes a feature extraction module and a feature pyramid fusion module. The feature extraction module first extracts feature images from several high-definition images and transmits the extracted feature images to the feature pyramid fusion module. The feature pyramid fusion module samples the feature images layer by layer and fuses them with the feature images of the previous layer to generate a report.

5. The flight inspection method according to claim 1, characterized in that, The inspection drone is equipped with an outer protective cover to prevent collisions.

Images