Anti-falling hanging point automatic recovery method and system based on waypoint automatic acquisition route planning

Abstract

The invention provides an anti-falling hanging point automatic recovery method and system based on waypoint automatic collection route planning, and relates to the technical field of anti-falling hanging point loading.The method comprises the steps that flight state data of an unmanned aerial vehicle is obtained, space geographical coordinate information in the process that the unmanned aerial vehicle loads an anti-falling hanging point is recorded based on an automatic dotting form, route information is formed, and the unmanned aerial vehicle is automatically recovered; in the unmanned aerial vehicle recovery process, the control host can control the unmanned aerial vehicle to automatically recover the anti-falling hanging point, so that an operator only needs to pay attention to the flight and loading process when operating the unmanned aerial vehicle to hang the anti-falling hanging point to the iron tower angle steel, and does not need to carry out complex manual dotting interaction with a remote controller; the unmanned aerial vehicle can be controlled to be executed through one key in the recovery process, the learning cost of operators is greatly reduced, unmanned aerial vehicle explosion accidents can be effectively reduced, and the operation safety is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of anti-fall hanging point loading, and in particular to an anti-fall hanging point automatic recovery method and system based on waypoint automatic collection route planning. Background Art

[0002] In the power system, especially during the operation and maintenance of transmission line towers, high-altitude operations are frequent and dangerous. Traditional anti-fall devices have many shortcomings, such as the need for manual pre-installation on the tower, high costs, and difficulty in maintenance. These problems not only increase the safety risks of operators, but also affect the efficiency of operations. In order to solve these problems, drone tower anti-fall devices came into being. The drone carries the anti-fall device to the top of the tower, and uses the gravity self-locking mechanical mechanism to achieve rapid installation and removal of the device on the top of the tower, thereby greatly improving the safety and efficiency of the operation.

[0003] However, as an innovative technology, the anti-fall device on drones is very challenging for operators to be familiar with drones during actual application. It requires high technical skills and experience. It is very difficult for novices or operators with low technical skills, especially for the recovery process of the anti-fall device, which can easily lead to the occurrence of a bombing accident. Specifically, the anti-fall device on drones has the following disadvantages: 1. Both the take-off and recovery process of the drone require human participation, which is inefficient; 2. The drone needs to be remotely controlled for precise flight, which places high demands on the drone operator. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide an automatic recovery method and system for anti-fall hanging points based on automatic waypoint collection and route planning, so as to solve the problems of low recovery efficiency of manually remote-controlled drones and high operating requirements for operators when recovering anti-fall hanging points.

[0005] To achieve the above-mentioned purpose, the first aspect of the present invention provides a method for automatically recovering anti-fall hanging points based on waypoint automatic collection route planning, the method comprising:

[0006] S101. The operator controls the UAV with the control host and anti-fall hanging point to fly to the top of the target tower through the remote control;

[0007] S102, the control host obtains the flight status data of the drone in real time when the drone flies to the top of the target tower, wherein the flight status data of the drone includes the flight speed, coordinates and image transmission information;

[0008] S103, the control host automatically performs point marking according to the flight speed and coordinates of the drone during the drone's ascent, and obtains speed point marking information;

[0009] S104, the control host recognizes the image transmission picture information, and when the tower top is recognized, automatically marks a point based on the coordinates at that time to obtain the tower top marking information;

[0010] S105, the control host recognizes the image transmission picture information, and when a single angle steel is recognized, automatically marks points based on the coordinates at that time to obtain the angle steel marking information;

[0011] S106. When the operator remotely controls the drone to complete the loading of the anti-fall hanging point, a dot is made at a specified height directly above the drone to obtain downward dot information, and the dot information at the anti-fall hanging point loading location is recorded to obtain the hanging point dot information;

[0012] S107, integrating the speed marking information, the tower top marking information, the angle steel marking information, the downward detection marking information and the hanging point marking information to form route information, and controlling the UAV to automatically recover the anti-fall hanging point according to the route information.

[0013] Furthermore, step S103 specifically includes the following operations:

[0014] S201, the control host determines whether the flight speed of the drone is lower than a preset speed threshold during the ascending process of the drone. If so, the spatial geographic coordinate position (x, y, z) of the drone at this time is obtained as speed point information, which is recorded as Ni;

[0015] S202, if the flight speed of the drone continues to be lower than the preset speed threshold, the dot information is recorded only once, and the cache is set to no longer record the dot information;

[0016] S203: If the flying speed of the drone is higher than the preset threshold again, the cache is set to be capable of recording dot information, and S201 is executed again. When recording new dot information, i=i+1.

[0017] Furthermore, before step S104, the following operations are performed:

[0018] S301, sampling images directly above the top of the tower, annotating the sampled image data, and obtaining an annotated data set;

[0019] S302. Train the standard neural network model by annotating the data set to obtain a tower top target detection and recognition model. Compress the tower top target detection and recognition model and deploy it to the control host. Provide an SDK interface for the tower top target detection and recognition model for the control host to call. The return results of the tower top target detection and recognition model include tower and notower.

[0020] Furthermore, step S104 specifically includes the following operations:

[0021] S401, the control host inputs the image transmission picture information into the tower top target detection and recognition model through the SDK interface for recognition;

[0022] S402, the control host determines whether the return result of the tower top target detection and recognition model is tower. If so, the automatic marking based on the flight speed and coordinates of the drone is immediately interrupted, and the automatic marking based on the tower top recognition result is performed to obtain the tower top marking information.

[0023] Furthermore, step S105 specifically includes the following operations:

[0024] S501, the control host performs grayscale processing on the image transmission image information, inputs the grayscale processing result into the line extraction algorithm, and identifies the single angle steel on the top of the tower;

[0025] S502: When a single angle steel is identified, the angle steel is automatically marked according to the coordinate information of the drone at that time to obtain the angle steel marking information.

[0026] Furthermore, in step S106, the marking is performed at a specified height directly above the UAV. Specifically, after the operator remotely controls the UAV to complete the loading of the anti-fall hanging point, automatic marking is performed at 0.3 meters directly above the UAV to obtain downward marking information.

[0027] The second aspect of the present invention provides an automatic recovery system for anti-fall hanging points based on automatic waypoint collection and route planning, the system includes a remote control, an unmanned aerial vehicle, a control host and an anti-fall hanging point, the remote control communicates with the unmanned aerial vehicle via microwave radio frequency, the unmanned aerial vehicle includes a downward-looking camera and a forward-looking camera, the control host is arranged on the unmanned aerial vehicle, and performs two-way interactive communication with the unmanned aerial vehicle via a serial port, the control host includes a locking tongue device, the locking tongue device and the anti-fall hanging point form a detachable structure, the control host is used to control the locking tongue device to perform a detaching action to release the anti-fall hanging point, or to perform a locking action to grab the anti-fall hanging point, and the system is used to execute the method described in the first aspect above.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] The present invention provides an automatic recovery method and system for anti-fall hanging points based on automatic route planning with waypoint collection. The method obtains the flight status data of the UAV, records the spatial geographic coordinate information of the UAV during the process of loading the anti-fall hanging points in the form of automatic marking, and forms route information. During the recovery of the UAV, the control host can control the UAV to automatically recover the anti-fall hanging points, so that the operating personnel only need to pay attention to the flight and loading process when operating the UAV to mount the anti-fall hanging points to the angle steel of the tower, without the need for complex manual marking interaction with the remote control. The recovery process can be controlled by the UAV with one click, which greatly reduces the learning cost of the operating personnel, can effectively reduce the occurrence of UAV explosion accidents, and improve operational safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only preferred embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0031] Figure 1 It is a schematic diagram of the overall process of an automatic recovery method for anti-fall hanging points based on automatic waypoint collection and route planning provided by an embodiment of the present invention.

[0032] Figure 2 It is a schematic diagram of the automatic recovery process of the anti-fall hanging point provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The enumerated embodiments are only used to explain the present invention and are not used to limit the scope of the present invention.

[0034] The technical problem to be solved by the present invention is to overcome the problems of low safety, low efficiency, great difficulty in operation, great difficulty in management, and long training time in the existing technology of anti-fall hanging points for UAVs, and to provide an automatic recovery method and system for anti-fall hanging points mounted on towers for UAVs based on spatial geographic coordinate positioning, thereby improving the practicality of UAVs, improving operating efficiency, and reducing safety risks and inspection costs.

[0035] Reference Figure 1 One embodiment of the present invention provides a method for automatically recovering anti-fall hanging points based on waypoint automatic collection route planning, the method comprising the following steps:

[0036] S101. The operator uses a remote control to fly the drone carrying the control host and anti-fall hanging point to the top of the target tower.

[0037] In this embodiment, the operator is an operator who operates the drone to load the anti-fall hanging point. The anti-fall hanging point is a device mounted above the angle steel for use in climbing the tower. The safety rope can be attached to the tower based on the device and the bottom fixing device to form a flexible guide rail from top to bottom. The operator is connected to the safety rope through a safety belt to perform operations.

[0038] In this embodiment, the remote controller is a ground station device, which remotely controls the drone by microwave radio frequency communication. The drone has a downward-looking camera and a forward-looking camera, so that it can collect images below and in front of it. The control host is the main carrier for implementing the method provided in this embodiment. The control host is integrated with the drone in terms of the hardware interface, and communicates with the drone in a two-way interactive manner through the serial port. At the same time, it forms a detachable structural device with the anti-fall hanging point, which contains a locking tongue device. The control host can control the locking tongue to release and lock the anti-fall hanging point by receiving remote control commands, thereby releasing the anti-fall hanging point or grabbing the anti-fall hanging point. The control host has AI computing power and edge computing capabilities, and can obtain drone information in real time.

[0039] S102. The control host obtains the UAV flight status data in real time when the UAV flies to the top of the target tower. The UAV flight status data includes flight speed, coordinates and image transmission information.

[0040] S103, the control host automatically performs point marking according to the flight speed and coordinates of the drone during the drone's ascent, and obtains speed point marking information.

[0041] S104, the control host recognizes the image transmission picture information, and when the tower top is recognized, automatically marks a point based on the coordinates at that time to obtain the tower top marking information.

[0042] S105, the control host identifies the image transmission information, and when a single angle steel is identified, automatically marks points based on the coordinates at that time to obtain the angle steel marking information.

[0043] S106. When the operator remotely controls the UAV to complete the loading of the anti-fall hanging point, a dot is made at a specified height directly above the UAV to obtain downward dot information, and the dot information at the anti-fall hanging point loading location is recorded to obtain the hanging point dot information.

[0044] S107, integrating the speed marking information, the tower top marking information, the angle steel marking information, the downward detection marking information and the hanging point marking information to form route information, and controlling the UAV to automatically recover the anti-fall hanging point according to the route information.

[0045] In this embodiment, the dotting is to obtain the absolute value (x, y, z) of the spatial geographic coordinates of the drone, where x represents accuracy, y represents dimension, and z represents height. The control host saves the information to complete a dotting action. During the normal operation of the drone to load the anti-fall hanging point, the operator will mainly experience three actions: first, there may be obstacles in front. At this time, the operator needs to control the drone to slow down or stop moving forward to make further judgments; second, control the drone to fly to a higher place above the top of the tower to avoid bombing; third, finely operate the drone to mount the anti-fall hanging point on the angle steel of the tower. At this time, the angle steel frame under the perspective of the drone camera accounts for a large proportion. For conventional drones, only a single angle steel can be photographed one meter above the top of the tower, which is the final angle steel target for each anti-fall hanging point. This embodiment performs a complete automatic dotting process based on the above three action stages.

[0046] As a possible implementation, step S103 specifically includes the following operations:

[0047] S201. During the ascent of the UAV, the control host determines whether the flight speed of the UAV is lower than a preset speed threshold. If so, the spatial geographic coordinate position (x, y, z) of the UAV at this time is obtained as speed point information, which is recorded as Ni.

[0048] S202: If the flying speed of the drone continues to be lower than the preset speed threshold, the dot information is recorded only once, and the cache is set to no longer record the dot information.

[0049] S203: If the flying speed of the drone is higher than the preset threshold again, the cache is set to be capable of recording dot information, and S201 is executed again. When recording new dot information, i=i+1.

[0050] For example, a speed threshold of 1 m / s can be set in the control host in advance. The preset speed threshold is used to automatically mark points based on the flight speed information of the drone when the control host does not identify the tower top. When the operator is operating the drone to fly normally, if an obstacle is encountered, the flight will slow down or stop. The control host obtains the spatial geographic coordinate position and flight speed data of the drone in real time. When the flight speed of the drone is lower than the preset speed threshold of 1 m / s for the first time, it marks points and records them as N1 (x, y, z). As long as the flight speed of the drone is lower than the preset speed threshold at this time, the control host only records the marking information once. And when the flight speed of the drone continues to be lower than the preset speed threshold, the cache is set to no longer record the marking information. When the operator remotely controls the drone to avoid obstacles and the drone resumes flying at a faster speed, if its flight speed is higher than the preset speed threshold, the cache is set to be able to mark points, and this is a cycle, and the marking information N2, N3, N4.....Na is recorded in sequence, and Na represents the last marking information recorded based on the speed information.

[0051] As a possible implementation, before step S104, the following operations are performed:

[0052] S301. Sample images directly above the top of the tower, and annotate the sampled image data to obtain an annotated data set.

[0053] S302, train the standard neural network model through the labeled data set to obtain the tower top target detection and recognition model, compress the tower top target detection and recognition model and deploy it to the control host, and provide the SDK interface of the tower top target detection and recognition model for the control host to call, and the return results of the tower top target detection and recognition model include tower and notower. Tower means that the tower top target is recognized; notower means that the tower top target is not recognized.

[0054] Exemplarily, the standard neural network may be ResNet, or DenseNet, or ConvNeXt, or other types of neural networks. Those skilled in the art may select neural network models with different structures according to actual needs, and this embodiment does not specifically limit this.

[0055] Based on the above implementation, step S104 specifically includes the following operations:

[0056] S401, the control host inputs the image transmission picture information into the tower top target detection and recognition model through the SDK interface for recognition.

[0057] S402, the control host determines whether the return result of the tower top target detection and recognition model is tower. If so, the automatic marking based on the flight speed and coordinates of the drone is immediately interrupted, and the automatic marking based on the tower top recognition result is performed to obtain the tower top marking information.

[0058] The purpose of step S104 is to meet the requirement of S103 for dot accuracy. Since S104 is an AI target detection combined with vision, its dot record information will be more accurate. For example, the tower top dot information obtained in S104 can be recorded as Nb.

[0059] As another possible implementation, step S105 specifically includes the following operations:

[0060] S501. The control host performs grayscale processing on the image transmission information, inputs the grayscale processing result into the line extraction algorithm, and identifies the single angle steel on the top of the tower.

[0061] S502: When a single angle steel is identified, the angle steel is automatically marked according to the coordinate information of the drone at that time to obtain the angle steel marking information.

[0062] In this embodiment, after the control host identifies the angle steel, it immediately stops the automatic dotting based on the tower top recognition result, and switches to further dotting based on the single angle steel recognition and the anti-fall hanging point. Exemplarily, when the drone flies to 0.5 meters above the tower top, the operator needs to carefully control the flight of the drone so that the drone can hang the anti-fall hanging point on the angle steel. At this time, there will only be a single angle steel in the image transmission picture collected by the drone's camera. After identifying a single angle steel through grayscale processing and line extraction algorithms, the control host immediately performs automatic dotting based on the spatial geographic coordinate information of the drone at this time, obtains the angle steel dotting information, and records it as Nc. The combination of Nb and Nc forms a coherent route that is obstacle-free in the middle and is above a single angle steel. When the operator controls the drone to complete the mounting of the anti-fall hanging point on the angle steel, considering that the drone needs to move to a fixed point first and then probe to a specified position during the recovery process, the downward probe position is 0.3 meters, so at this time, an automatic dot is made at 0.3 meters above the drone and recorded as the downward probe dot information Nd, and the point corresponding to Nd is the second to last point. At the same time, the control host records the dot information Ne of the position where the drone mounts the anti-fall hanging point, which is used when the drone finally recovers the anti-fall hanging point and grabs the hanging point at the stopping position of the downward probe.

[0063] In step S107, the dot information of N1, N2, N3, ..., Na, Nb, Nc, Nd, and Ne formed after the above operations is integrated to form the route information, so that the drone can be controlled to automatically recover the anti-fall hanging point according to the route information.

[0064] In this embodiment, the process of recovering the anti-fall hanging point is the same as the process of the UAV flying to the angle steel to mount the anti-fall hanging point. The entire recovery flight process is completed automatically, and the control host automatically controls the UAV flight according to the route information.

[0065] As a possible implementation method, refer to Figure 2 The automatic recovery of the anti-fall hanging point includes the following operations:

[0066] S601. According to the route information, control the UAV to accelerate to a speed of 2 m / s and fly to point N1.

[0067] S602, control the UAV to maintain a speed of 2 m / s and fly to points N2, N3, ..., Nb in sequence.

[0068] S603, control the UAV to descend from point Nb to point Nc at a speed of 0.25 m / s.

[0069] S604, control the drone to pass through point Ne at a speed of 0.15 m / s and finally descend to point Ne. At this time, the anti-fall hanging point and the locking tongue device are locked, and the drone is controlled to fly upward away from the angle steel.

[0070] S605. Manually take over the drone and control it to fly back to the ground.

[0071] The method provided in this embodiment does not require any human intervention during the recovery process of the anti-fall hanging point, except for the operator controlling the drone loaded with the anti-fall hanging point to fly back to the ground at the end. This can greatly improve the efficiency of the recovery of the anti-fall hanging point and reduce dependence on personnel and technical requirements.

[0072] Another embodiment of the present invention provides an automatic recovery system for anti-fall hanging points based on automatic waypoint collection and route planning, the system includes a remote control, a drone, a control host and an anti-fall hanging point, the remote control communicates with the drone via microwave radio frequency, the drone includes a downward-looking camera and a forward-looking camera, the control host is arranged on the drone, and performs two-way interactive communication with the drone via a serial port, the control host includes a locking tongue device, the locking tongue device and the anti-fall hanging point form a detachable structure, the control host is used to control the locking tongue device to perform a detaching action to release the anti-fall hanging point, or to perform a locking action to grab the anti-fall hanging point, and the system is used to execute the method described in any of the aforementioned method embodiments.

[0073] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for automatically recovering anti-fall hanging points based on waypoint automatic collection route planning, characterized in that: The method comprises: S101. The operator controls the UAV with the control host and anti-fall hanging point to fly to the top of the target tower through the remote control; S102, the control host obtains the flight status data of the drone in real time when the drone flies to the top of the target tower, wherein the flight status data of the drone includes the flight speed, coordinates and image transmission information; S103, the control host automatically performs point marking according to the flight speed and coordinates of the drone during the drone's ascent, and obtains speed point marking information; S104, the control host recognizes the image transmission picture information, and when the tower top is recognized, automatically marks a point based on the coordinates at that time to obtain the tower top marking information; S105, the control host recognizes the image transmission picture information, and when a single angle steel is recognized, automatically marks points based on the coordinates at that time to obtain the angle steel marking information; S106. When the operator remotely controls the drone to complete the loading of the anti-fall hanging point, a dot is made at a specified height directly above the drone to obtain downward dot information, and the dot information at the anti-fall hanging point loading location is recorded to obtain the hanging point dot information; S107, integrating the speed marking information, the tower top marking information, the angle steel marking information, the downward detection marking information and the hanging point marking information to form route information, and controlling the UAV to automatically recover the anti-fall hanging point according to the route information.

2. According to claim 1, a method for automatically recovering anti-fall hanging points based on automatic waypoint collection route planning is characterized in that: Step S103 specifically includes the following operations: S201, the control host determines whether the flight speed of the drone is lower than a preset speed threshold during the ascending process of the drone. If so, the spatial geographic coordinate position (x, y, z) of the drone at this time is obtained as speed point information, which is recorded as Ni; S202, if the flight speed of the drone continues to be lower than the preset speed threshold, the dot information is recorded only once, and the cache is set to no longer record the dot information; S203: If the flying speed of the drone is higher than the preset threshold again, the cache is set to be capable of recording dot information, and S201 is executed again. When recording new dot information, i=i+1.

3. The method for automatically recovering anti-fall hanging points based on automatic waypoint collection and route planning according to claim 1 is characterized in that: Before step S104, perform the following operations: S301, sampling images directly above the top of the tower, annotating the sampled image data, and obtaining an annotated data set; S302. Train the standard neural network model by annotating the data set to obtain a tower top target detection and recognition model. Compress the tower top target detection and recognition model and deploy it to the control host. Provide an SDK interface for the tower top target detection and recognition model for the control host to call. The return results of the tower top target detection and recognition model include tower and notower.

4. The method for automatically recovering anti-fall hanging points based on automatic waypoint collection and route planning according to claim 3 is characterized in that: Step S104 specifically includes the following operations: S401, the control host inputs the image transmission picture information into the tower top target detection and recognition model through the SDK interface for recognition; S402, the control host determines whether the return result of the tower top target detection and recognition model is tower. If so, the automatic marking based on the flight speed and coordinates of the drone is immediately interrupted, and the automatic marking based on the tower top recognition result is performed to obtain the tower top marking information.

5. The method for automatically recovering anti-fall hanging points based on automatic waypoint collection and route planning according to claim 1 is characterized in that: Step S105 specifically includes the following operations: S501, the control host performs grayscale processing on the image transmission information, inputs the grayscale processing result into the line extraction algorithm, and identifies the single angle steel on the top of the tower; S502: When a single angle steel is identified, the angle steel is automatically marked according to the coordinate information of the drone at that time to obtain the angle steel marking information.

6. The method for automatically recovering anti-fall hanging points based on automatic waypoint collection and route planning according to claim 1 is characterized in that: In step S106, the marking is performed at a specified height directly above the UAV. Specifically, after the operator remotely controls the UAV to complete the loading of the anti-fall hanging point, automatic marking is performed at 0.3 meters directly above the UAV to obtain downward marking information.

7. An automatic recovery system for anti-fall hanging points based on automatic waypoint collection and route planning, characterized in that: The system includes a remote controller, a drone, a control host and an anti-fall hanging point. The remote controller communicates with the drone via microwave radio frequency. The drone includes a downward-looking camera and a forward-looking camera. The control host is arranged on the drone and performs two-way interactive communication with the drone via a serial port. The control host includes a locking tongue device. The locking tongue device and the anti-fall hanging point form a detachable structure. The control host is used to control the locking tongue device to perform a detaching action to release the anti-fall hanging point, or to perform a locking action to grab the anti-fall hanging point. The system is used to execute the method described in any one of claims 1 to 6.

Images