Unmanned aerial vehicle inspection method and system suitable for area without network coverage

Abstract

The invention discloses an unmanned aerial vehicle inspection method and system suitable for a non-network coverage area, relates to the technical field of unmanned aerial vehicle positioning, and solves the problem that an unmanned aerial vehicle cannot realize unmanned aerial vehicle inspection operation in complex scenes such as a remote area, a non-network coverage area and a network coverage interrupted area. When the system is connected with a mobile communication network, initial position information is transmitted to a ground-based network RTK position calculation platform, and calculated ground-based differential data which is calculated and returned by the platform are received, and the positioning information of the unmanned aerial vehicle is calculated according to the ground-based differential data; and when the mobile communication network cannot be connected, the positioning information can be obtained through three means and can be selected based on a certain judgment reference, so that the accuracy of the positioning information is ensured. In a word, near-real-time precise positioning of the unmanned aerial vehicle in the whole area where the unmanned aerial vehicle can be connected with the mobile communication network or cannot be connected with the mobile communication network can be achieved, and the problem that normal inspection operation is difficult to conduct in the complex environment is solved.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) positioning technology, and in particular to an unmanned aerial vehicle (UAV) inspection method and system suitable for areas without network coverage. Background Art

[0002] Unmanned aerial vehicles, or UAVs for short, are unmanned aircraft that are operated by radio remote control equipment and self-contained program control devices. By importing the route-setting method, the UAV can fly automatically from station to station, and automatically classify and rename photos, update the three-dimensional space ledger, and quickly help power grid personnel accurately locate defects or fault points, so that the UAV line patrol work can move from "manual patrol" to "automatic patrol". However, when UAVs perform automatic tasks, they are prone to appear in complex scenarios such as remote areas, areas without network coverage, areas with intermittent network coverage, and areas with good operator network coverage, which leads to problems such as weak UAV signals, lack of accuracy, and large operation errors, which reduce the quality of operations and create flight safety hazards. For areas without signal coverage, manned patrols are currently the main method, and autonomous patrol solutions based on high-precision positioning technology services for UAVs cannot be carried out. Research on UHV transmission line patrol technology to solve this regional environment and realize the full range of automatic patrol work by UAVs will be an important application scenario for smart power transmission in the future.

[0003] In view of this, a drone inspection method and system suitable for areas without network coverage is needed. Summary of the invention

[0004] In view of the problem in the prior art that drones cannot perform drone inspection operations in complex scenarios such as remote areas, areas without network coverage, and areas with intermittent network coverage, the present invention provides a drone inspection method and system that is suitable for areas without network coverage, which can enable drones to obtain near-real-time precision positioning, speed measurement, and timing results in the entire area, and can obtain real-time centimeter-level positioning, speed measurement, and timing results in ground-based enhanced areas, integrating satellite-based RTK broadcasts with communication signal relay stations, operator mobile Internet, and other means, and can seamlessly switch between ground-based enhanced areas and non-ground-based enhanced areas, solving the problem of drone inspection operations in complex scenarios such as remote areas without network coverage, areas with intermittent network coverage, and areas with good operator network coverage. The specific technical solution is as follows:

[0005] A drone inspection method adapted to areas without network coverage includes the following steps:

[0006] Configure the UAV with a solution engine and an inertial navigation system for solving PPP-RTK data;

[0007] Obtain the coordinates of the towers and the topology of the tower lines in the inspection area, and generate inspection routes based on them. The drones then perform inspections based on the inspection routes.

[0008] During the inspection process, the drone acquires the mobile communication network signal in real time and determines the connection status with the mobile communication network;

[0009] When the UAV can be connected to the mobile communication network, the UAV obtains the current preliminary position and transmits it to the ground-based network RTK position solving platform through the mobile communication network. The ground-based network RTK position solving platform solves the preliminary position information and transmits back the solved ground-based differential data. The positioning information of the UAV is calculated based on the ground-based differential data.

[0010] When the UAV cannot be connected to the mobile communication network, PPP-RTK data is received via satellite and the first network-free positioning information is calculated based on the data. The inertial navigation system is turned on to obtain the second network-free positioning information, and one of the first network-free positioning information and the second network-free positioning information is selected as the UAV's positioning information.

[0011] Preferably, the process of selecting one of the first non-network positioning information and the second non-network positioning information as the positioning information of the drone is as follows:

[0012] The first non-network positioning information and the second non-network positioning information are set as two selection modes of positioning information of the drone, and one of them is set as an initial selection mode;

[0013] Regularly use the camera on the drone to synchronously obtain images below the drone;

[0014] Identify whether there are poles in the image based on the trained pole detection model;

[0015] Setting a mode switching threshold, wherein the mode switching threshold is the number of towers that are allowed to not exist in the image within a set period;

[0016] Collecting tower recognition conditions of a number of images within a set period, wherein the tower recognition conditions include the number of images in which towers are present and the number of images in which towers are not present;

[0017] The number of images without pole towers is compared with the mode switching threshold. If the number of images without pole towers is greater than the mode switching threshold, the selection mode is switched.

[0018] Preferably, the UAV is also provided with a satellite-ground integrated board. When the UAV cannot be connected to the mobile communication network, the satellite-based differential data obtained by the satellite-based solution platform is directly acquired through the satellite-based PPP-AR service and the one-way transmission link of the geosynchronous orbit satellite through the built-in satellite-ground integrated board, and then the positioning information of the electric UAV flight is calculated according to the satellite-based differential data obtained by the satellite-ground integrated board.

[0019] Preferably, the positioning information calculated according to the satellite-based differential data obtained from the satellite-ground integrated board is used as the third network-free positioning information. When the UAV cannot be connected to the mobile communication network, one of the first network-free positioning information, the second network-free positioning information and the third network-free positioning information is selected as the positioning information of the UAV.

[0020] Preferably, the process of selecting one of the first non-network positioning information, the second non-network positioning information and the third non-network positioning information as the positioning information of the drone is as follows:

[0021] The first non-network positioning information, the second non-network positioning information and the third non-network positioning information are set as three selection modes of positioning information of the drone, and one of them is set as the initial selection mode;

[0022] Regularly use the camera on the drone to synchronously obtain images below the drone, and identify whether there are poles in the image based on the trained pole detection model;

[0023] Setting a mode switching threshold, wherein the mode switching threshold is the number of towers that are allowed to not exist in the image within a set period;

[0024] Collecting the tower recognition conditions of a number of images within a set period, wherein the tower recognition conditions include the number of images with and without towers, comparing the number of images without towers with a mode switching threshold, and switching the selection mode if the number of images without towers is greater than the mode switching threshold;

[0025] The tower recognition conditions of several images within a set period are continuously collected, and the number of images without towers is compared with the mode switching threshold. If the number of images without towers is greater than the mode switching threshold, the selection mode continues to be switched.

[0026] Preferably, the switching order of the switching selection mode is: switching from selecting the first network-free positioning information to selecting the second network-free positioning information, switching from selecting the second network-free positioning information to selecting the third network-free positioning information, and switching from selecting the third network-free positioning information to selecting the first network-free positioning information.

[0027] Preferably, the pole tower detection model is specifically a pole tower detection model based on YOLO, and the training process is as follows:

[0028] Collect and label the images below the drone taken during the flight of the drone, and use the labeled images below the drone as training samples;

[0029] During training, YOLO adopts a multi-scale training strategy, and the image size in each batch is randomly selected to increase the model's ability to detect objects of different scales;

[0030] After training, YOLO can directly detect objects in the input image and output the location and category information of each detected object.

[0031] Based on the category information of the detected object output by YOLO, it is determined whether there is a pole tower. If there is a pole tower, the final output is the result that the pole tower exists in the image. If not, the final output is the result that the pole tower does not exist.

[0032] A drone inspection system adapted to areas without network coverage, applied to the above method, comprises:

[0033] Image acquisition module, used to collect image data during the drone inspection process;

[0034] Communication module, used to communicate with the ground-based network RTK position solution platform and other terminal platforms to complete data reception and upload operations;

[0035] The non-network data acquisition module includes a PPP-RTK data receiving module, a PPP-RTK data solving engine and an inertial navigation system. After receiving the PPP-RTK data, the PPP-RTK data receiving module solves the first non-network positioning information through the PPP-RTK data solving engine. The inertial navigation system is used to obtain the second non-network positioning information.

[0036] The data verification module includes a data acquisition module, an image recognition module and a data comparison module. The data acquisition module is used to collect image data from the image acquisition module and perform preprocessing operations. The image recognition module has a built-in pole tower detection model to identify whether there are pole towers in the image. The data comparison module is used to compare whether the number of images without pole towers in a certain period exceeds a threshold value.

[0037] The data selection switching module switches the location information acquisition source when the number of images without pole towers in a period exceeds a threshold, wherein the location information acquisition source includes the first no-network location information and the second no-network location information.

[0038] A computer-readable storage medium includes a stored program, wherein when the program is run, the device where the computer-readable storage medium is located is controlled to execute the drone inspection method adapted to areas without network coverage as described above.

[0039] A processor is used to run a program, wherein when the program is run, the drone inspection method adapted to areas without network coverage as described above is executed.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. The present invention integrates satellite-based broadcast RTK with communication signal relay stations and the operator's mobile Internet. When the drone can be connected to the mobile communication network, the drone obtains the current preliminary position and transmits it to the ground-based network RTK position solving platform through the mobile communication network. The ground-based network RTK position solving platform solves the preliminary position information and transmits back the solved ground-based differential data. The positioning information of the drone is calculated based on the ground-based differential data. In other words, in the area with the network, the present invention integrates the network RTK technology to obtain real-time centimeter-level positioning, speed measurement and timing results in the ground-based enhanced area.

[0042] 2. The present invention takes into account the problem of no network coverage in remote areas. When the UAV cannot be connected to the mobile communication network, the present invention can obtain positioning information through three means: one is to receive PPP-RTK data through satellites and calculate the positioning information based on this; the second is to start the inertial navigation system to obtain positioning information; the third is to calculate the positioning information of the electric UAV flight based on the satellite-ground integrated board based on the satellite-based differential data. Three types of positioning information can be obtained through the three means for selection.

[0043] 3. In an environment without network coverage, the present invention provides three means to obtain three sets of positioning information, and sets up verification means to select positioning information according to certain judgment criteria, further ensuring the accuracy of positioning information in an environment without network. Among them, the verification means is specifically to identify whether there are poles and towers in the images obtained within the period. The reason for selecting poles and towers is that in the process of drone inspection, the object with the best identification degree in the pictures taken is the pole and tower. Therefore, identifying whether the pole and tower appear in the pictures taken by the drone is a significant sign to judge whether the drone is yawed. In addition, a threshold comparison link is added in the verification process, which gives a fault tolerance rate to the batch of pictures that are periodically received and judged, or gives the drone a flexible space for data judgment when it is located at the midpoint between two towers and may not be able to capture the tower. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0045] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0048] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0049] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0050] Before further explaining the embodiments of the present invention, the general concepts in the art involved in the embodiments are briefly described as follows:

[0051] Network RTK technology: A GPS network composed of multiple reference stations is used to estimate the GPS error model of a region and provide correction data for users in the network coverage area. However, what users receive is not the observation data of a real reference station, but the data of a virtual reference station and the correction data of a reference grid that is closer to their location. Network RTK technology uses the observation information of each CORS reference station and, based on the CORS network architecture, establishes an accurate differential information solution model to solve high-precision differential data, and then sends various differential correction numbers to users through wireless communication data links. Network RTK technology integrates Internet technology, wireless communication technology, computer network management technology and GPS positioning technology, and is the core support technology and solution of the CORS network service system. Its main features are as follows: (1) Expanding the service scope. Network RTK expands the service scope of RTK positioning, from the early single-station RTK coverage to the entire CORS network. (2) Realizing resource sharing. Reference stations share data through the communication system, and share reference station data and benchmarks through the Internet. (3) Unifying standards. Network RTK adopts a unified standard and a unified differential format to ensure good compatibility of the system. (4) Improve the differential calculation model. With the CORS network as the framework, the reference station data and the benchmark coordinates as the basis, the RTK calculation mathematical model is improved.

[0052] PPP-AR (Precise Point Positioning with Ambiguity Resolution) is a type of navigation and positioning technology characterized by state space. It uses satellite orbit, clock error, code deviation, phase deviation, and vertical ionospheric correction data products, follows the parameter estimation criteria of non-difference processing, and simultaneously estimates the non-difference ambiguity to calculate the actual user's high-precision position. It is an optimized and improved PPP positioning technology.

[0053] PPP-RTK (Precise Point Positioning-Real Time Kinematic) is a type of navigation and positioning technology characterized by state space. It uses the full correction data products of satellite orbit, clock error, code deviation, phase deviation, vertical ionosphere correction, regional oblique ionosphere correction, and regional tropospheric delay, and estimates the non-differential ambiguity at the same time according to the parameter estimation criteria of non-differential processing to calculate the actual high-precision position of the user. It is a type of non-differential RTK positioning technology.

[0054] In one embodiment of the present invention, a drone inspection method adapted to areas without network coverage is provided. Figure 1 As shown, the following steps are included:

[0055] S1: Configure the UAV with a solution engine and an inertial navigation system for solving PPP-RTK data;

[0056] S2: Obtain the coordinates of the towers and the topology of the tower lines in the inspection area, and generate an inspection route based on it. The drone performs inspection operations according to the inspection route.

[0057] S3: During the inspection process, the drone obtains the mobile communication network signal in real time and determines the connection status with the mobile communication network;

[0058] S4: When the UAV can be connected to the mobile communication network, the UAV obtains the current preliminary position and transmits it to the ground-based network RTK position solving platform through the mobile communication network. The ground-based network RTK position solving platform solves the preliminary position information and transmits back the solved ground-based differential data. The positioning information of the UAV is calculated based on the ground-based differential data.

[0059] S5: When the UAV cannot be connected to the mobile communication network, PPP-RTK data is received via satellite and the first network-free positioning information is calculated based on the data, and the inertial navigation system is turned on to obtain the second network-free positioning information, and one of the first network-free positioning information and the second network-free positioning information is selected as the positioning information of the UAV.

[0060] The process of selecting one of the first non-network positioning information and the second non-network positioning information as the positioning information of the drone is as follows:

[0061] S01: setting the first non-network positioning information and the second non-network positioning information as two selection modes of positioning information of the drone, and setting one of them as the initial selection mode;

[0062] S02: Regularly use the camera on the drone to synchronously obtain images below the drone;

[0063] S03: Identify whether there are poles in the image based on the trained pole detection model; the pole detection model is specifically a pole detection model based on YOLO, that is, the pole detection model used here is mainly built and trained using YOLO. YOLQ (You 0nly Look 0nce) is a target detection algorithm based on deep learning, and its purpose is to achieve real-time target detection through a single forward pass.

[0064] The target detection process of the YOLO algorithm is as follows:

[0065] 1. The input image is divided into S xS grids, each grid is responsible for predicting one or more targets. Typically, the value of S is 7x7 or 9x9.

[0066] 2. For each grid, encode its position and size information into a bounding box. The bounding box is usually represented by (xy, w, h), where (x, y) is the coordinate of the center point of the bounding box, and (w, h) is the width and height of the bounding box.

[0067] 3. Each bounding box predicts the target category and its confidence. The softmax function is usually used to classify the target category, and the confidence is used to evaluate whether there is an object in the bounding box.

[0068] 4. During training, YOLO uses the cross entropy loss function to optimize the classification accuracy of the target category and the accuracy of the bounding box position and size.

[0069] 5. During the prediction process, according to the given threshold, the bounding boxes with confidence higher than the threshold are screened out, and the non-maximum suppression algorithm is used to eliminate overlapping bounding boxes.

[0070] 6. Finally, by mapping the bounding box back to the original image space, we can get the location and size information of the object, as well as its category and confidence.

[0071] Therefore, step S03 is specifically as follows:

[0072] S031: collect and mark the images below the drone taken during the flight of the drone, and use the marked images below the drone as training samples;

[0073] S032: During training, YOLO adopts a multi-scale training strategy, and the image size in each batch is randomly selected to increase the model's ability to detect objects of different scales;

[0074] S033: After training, YOLO can directly detect objects in the input image and output the location and category information of each detected object;

[0075] S034: Determine whether there is a pole tower based on the category information of the detected object output by YOLO. If there is a pole tower, finally output the result that the pole tower exists in the image; if not, finally output the result that the pole tower does not exist.

[0076] S04: Setting a mode switching threshold, the mode switching threshold is the number of images that are allowed to not contain poles within a set period; the reason for setting the threshold is to give a tolerance rate to the batch of images that are periodically received and judged, or to give the drone a space for data judgment when it is located at a position where it may not be able to capture a pole, such as the midpoint between two poles. For example, among 40 photos taken within a period (such as ten minutes), 4 photos without poles can be allowed to exist, and this number 4 is the set threshold.

[0077] S05: collecting tower recognition information of a number of images within a set period, wherein the tower recognition information includes the number of images with towers and the number of images without towers;

[0078] S06: Compare the number of images without pole towers with the mode switching threshold. If the number of images without pole towers is greater than the mode switching threshold, switch the selection mode.

[0079] The reason for choosing poles and towers is that during the inspection process of drones, the object with the best identification in the pictures taken is the poles and towers. Therefore, identifying whether the poles and towers appear in the pictures taken by the drone is a significant sign to determine whether the drone has deviated. After steps S01-S06, the switching between the first network-free positioning information and the second network-free positioning information can be completed. When the first network-free positioning information is not accurate enough and causes the drone to deviate and cannot photograph the poles and towers, it can be switched to the second network-free positioning information to realize the positioning operation, and vice versa. By this means, the problem that the drone cannot remedy the deviation caused by using one positioning method for a long time without a network can be avoided.

[0080] In one embodiment of the present invention, the above-mentioned step S5 is added: a satellite-ground integrated board is built into the drone, and when the drone cannot be connected to the mobile communication network, the satellite-based differential data obtained by the satellite-based solution platform is directly obtained through the built-in satellite-ground integrated board using the satellite-based PPP-AR service and the geosynchronous orbit satellite unidirectional transmission link, and then the positioning information of the electric drone flight is calculated according to the satellite-based differential data obtained by the satellite-ground integrated board, and the positioning information calculated according to the satellite-based differential data obtained by the satellite-ground integrated board is used as the third network-free positioning information.

[0081] When the drone cannot connect to the mobile communication network, one of the first no-network positioning information, the second no-network positioning information and the third no-network positioning information is selected as the drone's positioning information. The process is as follows:

[0082] S001: setting the first non-network positioning information, the second non-network positioning information and the third non-network positioning information as three selection modes of positioning information of the drone, and setting one of them as the initial selection mode;

[0083] S002: Regularly use the camera on the drone to synchronously obtain images below the drone, and identify whether there are poles in the image based on the trained pole detection model;

[0084] S003: Setting a mode switching threshold, wherein the mode switching threshold is the number of towers that are allowed to not exist in the image within a set period;

[0085] S004: Collecting tower recognition conditions of a number of images within a set period, wherein the tower recognition conditions include the number of images with and without towers, comparing the number of images without towers with a mode switching threshold, and switching the selection mode if the number of images without towers is greater than the mode switching threshold;

[0086] S005: Continuously collect the tower recognition conditions of a number of images within a set period, compare the number of images without towers with the mode switching threshold, and if the number of images without towers is greater than the mode switching threshold, continue to switch the selection mode. The switching order of the switching selection mode is: switching from selecting the first no-network positioning information to selecting the second no-network positioning information, switching from selecting the second no-network positioning information to selecting the third no-network positioning information, and switching from selecting the third no-network positioning information to selecting the first no-network positioning information.

[0087] In one embodiment of the present invention, a drone inspection system adapted to areas without network coverage is provided, which is applied to the above method and includes:

[0088] Image acquisition module, used to collect image data during the drone inspection process;

[0089] The communication module is used to communicate with the ground-based network RTK position solution platform and other terminal platforms to complete data reception and upload operations.

[0090] The non-network data acquisition module includes a PPP-RTK data receiving module, a PPP-RTK data solving engine and an inertial navigation system. After receiving the PPP-RTK data, the PPP-RTK data receiving module solves the first non-network positioning information through the PPP-RTK data solving engine. The inertial navigation system is used to obtain the second non-network positioning information.

[0091] The data verification module includes a data acquisition module, an image recognition module and a data comparison module. The data acquisition module is used to collect image data from the image acquisition module and perform preprocessing operations. The image recognition module has a built-in pole tower detection model to identify whether there are pole towers in the image. The data comparison module is used to compare whether the number of images without pole towers in a certain period exceeds a threshold value.

[0092] The data selection switching module switches the location information acquisition source when the number of images without pole towers in a period exceeds a threshold, wherein the location information acquisition source includes the first no-network location information and the second no-network location information.

[0093] To sum up, the drone can obtain near real-time precise positioning, speed measurement and timing results in the entire area, and can obtain real-time centimeter-level positioning, speed measurement and timing results in the ground-based augmented area. It integrates satellite-based RTK broadcast with communication signal relay stations, operator mobile Internet and other means, and can seamlessly switch between the ground-based augmented area and the entire area without ground-based augmentation, solving the problem of realizing drone inspection operations in complex scenarios such as remote areas, areas without network coverage, areas with intermittent network coverage, and areas with good operator network coverage, and solves the problems raised in the background technology.

[0094] Those of ordinary skill in the art will appreciate that the units (or modules, steps, etc., the same below) of the examples described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0095] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0096] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0097] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nlyMemory), random access memory (RAM, RandomAccessMemory), mobile hard disk, magnetic disk or optical disk, etc., which can store program code.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A drone inspection method suitable for areas without network coverage, characterized in that: The following steps are involved: Pre-configure the UAV with an inertial navigation system and a solution engine for solving PPP-RTK data; Obtain the coordinates of the towers and the topology of the tower lines in the inspection area, and generate inspection routes based on them. The drones then perform inspections based on the inspection routes. During the inspection process, the drone acquires the mobile communication network signal in real time and determines the connection status with the mobile communication network; When the UAV can be connected to the mobile communication network, the UAV obtains the current preliminary position and transmits it to the ground-based network RTK position solving platform through the mobile communication network. The ground-based network RTK position solving platform solves the preliminary position information and transmits back the solved ground-based differential data. The positioning information of the UAV is calculated based on the ground-based differential data. When the UAV cannot be connected to the mobile communication network, PPP-RTK data is received via satellite and the first network-free positioning information is calculated based on the data. The inertial navigation system is turned on to obtain the second network-free positioning information, and one of the first network-free positioning information and the second network-free positioning information is selected as the UAV's positioning information.

2. The method for unmanned aerial vehicle inspection in areas without network coverage according to claim 1, characterized in that: The process of selecting one of the first no-network positioning information and the second no-network positioning information as the positioning information of the drone is as follows: The first non-network positioning information and the second non-network positioning information are set as two selection modes of positioning information of the drone, and one of them is set as an initial selection mode; Regularly use the camera on the drone to synchronously obtain images below the drone; Identify whether there are poles in the image based on the trained pole detection model; Setting a mode switching threshold, wherein the mode switching threshold is the number of towers that are allowed to not exist in the image within a set period; Collecting tower recognition conditions of a number of images within a set period, wherein the tower recognition conditions include the number of images in which towers are present and the number of images in which towers are not present; The number of images without pole towers is compared with the mode switching threshold. If the number of images without pole towers is greater than the mode switching threshold, the selection mode is switched.

3. The method for unmanned aerial vehicle inspection in areas without network coverage according to claim 1 is characterized in that: It also includes a built-in satellite-ground integrated board in the drone. When the drone cannot be connected to the mobile communication network, the satellite-based PPP-AR service and the one-way transmission link of the geosynchronous orbit satellite are used to directly obtain the satellite-based differential data obtained by the satellite-based solution platform through the built-in satellite-ground integrated board, and then the positioning information of the electric drone's flight is calculated based on the satellite-based differential data obtained by the satellite-ground integrated board.

4. The method for unmanned aerial vehicle inspection in areas without network coverage according to claim 3 is characterized in that: The positioning information calculated according to the satellite-based differential data obtained from the satellite-ground integrated board is used as the third network-free positioning information. When the UAV cannot be connected to the mobile communication network, one of the first network-free positioning information, the second network-free positioning information and the third network-free positioning information is selected as the UAV's positioning information.

5. The method for unmanned aerial vehicle inspection in areas without network coverage according to claim 4 is characterized in that: The process of selecting one of the first no-network positioning information, the second no-network positioning information, and the third no-network positioning information as the positioning information of the drone is as follows: The first non-network positioning information, the second non-network positioning information and the third non-network positioning information are set as three selection modes of positioning information of the drone, and one of them is set as the initial selection mode; Regularly use the camera on the drone to synchronously obtain images below the drone, and identify whether there are poles in the image based on the trained pole detection model; Setting a mode switching threshold, wherein the mode switching threshold is the number of towers that are allowed to not exist in the image within a set period; Collecting the tower recognition conditions of a number of images within a set period, wherein the tower recognition conditions include the number of images with and without towers, comparing the number of images without towers with a mode switching threshold, and switching the selection mode if the number of images without towers is greater than the mode switching threshold; The tower recognition conditions of several images within a set period are continuously collected, and the number of images without towers is compared with the mode switching threshold. If the number of images without towers is greater than the mode switching threshold, the selection mode continues to be switched.

6. The method for unmanned aerial vehicle inspection in areas without network coverage according to claim 5, characterized in that: The switching order of the switching selection mode is: switching from selecting the first non-network positioning information to selecting the second non-network positioning information, switching from selecting the second non-network positioning information to selecting the third non-network positioning information, and switching from selecting the third non-network positioning information to selecting the first non-network positioning information.

7. A drone inspection method adapted to areas without network coverage according to claim 2 or 4, characterized in that: The tower detection model is specifically a tower detection model based on YOLO, and the training process is as follows: Collect and label the images below the drone taken during the flight of the drone, and use the labeled images below the drone as training samples; During training, YOLO adopts a multi-scale training strategy, and the image size in each batch is randomly selected to increase the model's ability to detect objects of different scales; After training, YOLO can directly detect objects in the input image and output the location and category information of each detected object. Based on the category information of the detected object output by YOLO, it is determined whether there is a pole tower. If there is a pole tower, the final output is the result that the pole tower exists in the image. If not, the final output is the result that the pole tower does not exist.

8. A drone inspection system adapted to areas without network coverage, characterized in that: The method applied to any one of claims 1 to 7, comprising: Image acquisition module, used to collect image data during the drone inspection process; Communication module, used to communicate with the ground-based network RTK position solution platform and other terminal platforms to complete data reception and upload operations; The non-network data acquisition module includes a PPP-RTK data receiving module, a PPP-RTK data solving engine and an inertial navigation system. After receiving the PPP-RTK data, the PPP-RTK data receiving module solves the first non-network positioning information through the PPP-RTK data solving engine. The inertial navigation system is used to obtain the second non-network positioning information. The data verification module includes a data acquisition module, an image recognition module and a data comparison module. The data acquisition module is used to collect image data from the image acquisition module and perform preprocessing operations. The image recognition module has a built-in pole tower detection model to identify whether there are pole towers in the image. The data comparison module is used to compare whether the number of images without pole towers in a certain period exceeds a threshold value. The data selection switching module switches the location information acquisition source when the number of images without pole towers in a period exceeds a threshold, wherein the location information acquisition source includes the first no-network location information and the second no-network location information.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the drone inspection method adapted to areas without network coverage as described in any one of claims 1 to 7.

10. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the drone inspection method adapted to areas without network coverage as described in any one of claims 1 to 7.

Images