Wind turbine intelligent auxiliary hoisting system, method, device and storage medium

Abstract

The application discloses a kind of wind generating set intelligent auxiliary hoisting systems, methods, equipment and storage medium, the method includes obtaining hoisting operation field image, identify all detection targets in hoisting operation field image, and determine tracking target;Obtain the continuous video frame of tracking target, tracking target is tracked in real time according to continuous video frame, and the parameter and shooting angle of pan-tilt high-definition camera are adjusted, so that tracking target is always located in the center of each frame image;The spatial position of tracking target and its corresponding in-place target is obtained, and the real-time relative position information between tracking target and its corresponding in-place target is calculated;Real-time relative position information and corresponding frame image in continuous video frame are displayed to guide hoisting operation of tracking target.The application solves the problem of hoisting operation difficulty under complex working conditions, improves hoisting efficiency and in-place accuracy.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of engineering construction, and particularly relates to a wind turbine intelligent auxiliary hoisting system and method based on computer vision, an electronic device and a storage medium. BACKGROUND

[0002] The development and utilization of clean energy are increasingly valued by countries around the world. Wind power generation refers to the conversion of wind kinetic energy into electrical energy. Wind energy is a clean energy. Under the driving of global energy efficiency management and energy digital transformation, wind power construction is accelerating, and the demand for wind turbines is growing exponentially.

[0003] Currently, a single wind turbine hoisting operation usually requires more than 20 people to work together. In terms of efficiency, the crane driver cannot see the whole picture in the cab and cannot independently achieve rapid positioning, relying entirely on the scheduling and command of the ground commander. Especially for the third and fourth tower sections, the observation error of the ground commander is large due to the positioning height of more than 50m, resulting in low hoisting efficiency. At the same time, the positioning process needs to be completed in coordination with the commander inside the tower section, which is complicated and time-consuming. In terms of construction safety, the hoisting process cannot be monitored in real time for the safety of the workers and the collision of the hoisting target, and there is no warning for illegal construction behavior with safety hazards, which can easily lead to accidents. In terms of environmental conditions, current wind turbine hoisting can only be carried out during the day. At night, on cloudy days or in foggy weather, it is difficult to supplement light due to low visibility, and hoisting has problems such as positioning difficulty, high risk coefficient, etc.

[0004] Therefore, it is urgent to develop an intelligent auxiliary hoisting system for wind turbines to achieve positioning and guidance of wind turbine hoisting in complex working conditions such as night and fog, and to improve the efficiency, safety and environmental adaptability of wind turbine hoisting operations. SUMMARY

[0005] The present application aims to provide a wind turbine intelligent auxiliary hoisting system, method, device and storage medium to solve the problems of traditional wind turbine hoisting technology, such as dependence on commander scheduling and command, large observation error of the commander leading to low hoisting efficiency, complicated steps and long time consumption, inability to monitor in real time during hoisting, high risk coefficient, and inability to hoist in complex working conditions.

[0006] The present application solves the above technical problems through the following technical solutions: a wind turbine intelligent auxiliary hoisting system, comprising a gimbal high-definition camera, a positioning device, and a control console arranged in a hoisting tool; the gimbal high-definition camera and the positioning device are connected to the control console;

[0007] The gimbal high-definition camera is used to capture hoisting site images and track continuous video frames of the target;

[0008] The positioning device is used for positioning the spatial positions of the tracking target and the corresponding in-place target;

[0009] The control console is used for acquiring and visualizing the hoisting site image and the continuous video frames of the tracking target in real time, adjusting the parameters and shooting angles of the gimbal high-definition camera, making the hoisting site image clear and containing at least one target to be hoisted, and making the continuous video frames contain the tracking target and the tracking target always located at the center of each frame image; identifying all detected targets in the hoisting site image, determining the tracking target; tracking the tracking target in real time according to the continuous video frames; acquiring the spatial positions of the tracking target and the corresponding in-place target, calculating real-time relative position information according to the spatial positions of the tracking target and the corresponding in-place target, and displaying the real-time relative position information and the corresponding frame image in the continuous video frames to guide the hoisting operation of the tracking target.

[0010] Further, the positioning device comprises a UWB control module, M UWB base stations and a plurality of UWB tags, and M is greater than or equal to 3; the M UWB base stations are respectively connected with the UWB control module, and the UWB control module is connected with the control console;

[0011] The M UWB base stations are arranged at the periphery of the hoisting operation area and are located at the same height and are not collinear; the UWB tags are arranged on the cooperation surfaces of each target to be hoisted and the corresponding in-place target, each UWB tag has a unique ID, and the ID of the UWB tag on the target to be hoisted corresponds to the ID of the UWB tag on the in-place target matched with the target to be hoisted;

[0012] The UWB control module is used for calculating the distance between the tracking target and each UWB base station according to the communication time difference between the UWB tag on the tracking target and each UWB base station, and calculating the distance between the in-place target corresponding to the tracking target and each UWB base station according to the communication time difference between the UWB tag on the in-place target and each UWB base station;

[0013] The spatial positions of each UWB base station are acquired, and the calculated distances and the spatial positions of each UWB base station are sent to the control console, so that the control console calculates the spatial positions of the tracking target and the corresponding in-place target according to the distances and the spatial positions of each UWB base station.

[0014] Further, the spatial positions refer to spatial positions in a WGS coordinate system or a self-defined coordinate system;

[0015] The self-defined coordinate system is a three-dimensional coordinate system with a certain UWB base station as the origin and all UWB base stations located at the same height and not collinear.

[0016] Based on the same concept, the application also provides an intelligent auxiliary hoisting method for a wind turbine generator system, comprising the following steps:

[0017] Obtaining a hoisting operation site image, wherein the hoisting operation site image at least contains a target to be hoisted;

[0018] Identifying all detection targets in the hoisting operation site image, and determining a tracking target from the detection targets;

[0019] Obtaining continuous video frames of the tracking target, tracking the tracking target in real time according to the continuous video frames, and adjusting the parameters and shooting angle of the pan-tilt high-definition camera, so that the tracking target is always located at the center of each frame image;

[0020] Obtaining the spatial positions of the tracking target and its corresponding in-place target, and calculating the real-time relative position information between the tracking target and its corresponding in-place target according to the spatial positions of the tracking target and its corresponding in-place target;

[0021] Displaying the real-time relative position information and the corresponding frame image in the continuous video frames to guide the hoisting operation of the tracking target.

[0022] Further, the specific implementation process of identifying all detection targets in the hoisting operation site image is as follows:

[0023] Constructing a target recognition model and training the target recognition model;

[0024] Identifying all detection targets in the hoisting operation site image by using the trained target recognition model.

[0025] Further, the target recognition model is a YOLOV5 model.

[0026] Further, the target recognition model and the deepsort algorithm are used to track the tracking target in real time, and the specific implementation process is as follows:

[0027] Inputting each frame image of the continuous video frames into the target recognition model to obtain a rectangular frame of the detection target;

[0028] Taking the rectangular frame of the detection target as the input of the deepsort algorithm, and determining the tracking target in the corresponding frame image according to the rectangular frame of the detection target and the moving track of the detection target in the front and rear frame images by the deepsort algorithm;

[0029] Adjusting the shooting angle and parameters of the pan-tilt high-definition camera, so that the tracking target is located at the center of the corresponding frame image.

[0030] Further, the detection target below the hoisting tool hook is taken as the tracking target, or any detection target is manually selected as the tracking target.

[0031] Based on the same concept, the application provides an electronic device, comprising:

[0032] a memory for storing a computer program;

[0033] a processor for executing the computer program to realize the intelligent auxiliary hoisting method of the wind turbine generator set as described above.

[0034] Based on the same concept, the application provides a computer readable storage medium, the computer readable storage medium has a computer program stored thereon, and the computer program is executed by a processor to realize the intelligent auxiliary hoisting method of the wind turbine generator set as described above.

[0035] Advantages

[0036] Compared with the prior art, the application has the following advantages:

[0037] The application controls the gimbal high-definition camera to start the night vision or infrared mode and adjusts the focal length to capture clear hoisting site images and continuous video frames of the tracking target, which is beneficial to hoisting operation in complex working conditions such as night, rain, fog and snow, improves the environmental adaptability of the system, and solves the problem of difficult hoisting operation in complex working conditions.

[0038] The application sets the control console in the hoisting tool, provides the driver in the hoisting tool with a visual picture of the hoisting target (including the tracking target), and synchronously displays the relative position information between the tracking target and the in-place target, realizes the visualization of hoisting operation guidance, improves the guidance accuracy and hoisting efficiency, and saves a large amount of manpower and material resources; the whole hoisting operation process is visualized on the control console, realizes real-time supervision of the hoisting process, and improves the safety of the hoisting operation. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions of the application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only one embodiment of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0040] Figure 1 is a field layout diagram of the ultra-wideband real-time positioning network in the embodiment of the application;

[0041] Figure 2 is a schematic diagram of a custom coordinate system constructed by 4 UWB base stations in the embodiment of the application;

[0042] Figure 3 is a frame image in a continuous video frame in the embodiment of the application;

[0043] Figure 4 is a flow chart of the intelligent auxiliary hoisting method of the wind turbine generator set in the embodiment of the application;

[0044] Figure 5 is the artificial marking map of the sample picture in the embodiment of the application. DETAILED DESCRIPTION

[0045] The technical solutions in the application will be described clearly and completely in combination with the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the application.

[0046] The technical solutions of the application will be described in detail in combination with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes can not be described in some embodiments.

[0047] The intelligent auxiliary hoisting system of the wind turbine generator set provided by the embodiment of the application comprises a gimbal high-definition camera, a positioning device, and a console arranged in a hoisting tool; the gimbal high-definition camera and the positioning device are connected with the console respectively.

[0048] In a specific embodiment of the application, the gimbal high-definition camera comprises a support, a three-dimensional gimbal arranged on the support, and a high-definition camera arranged on the three-dimensional gimbal; the three-dimensional gimbal and the high-definition camera are connected with the console respectively. The three-dimensional gimbal rotates under the control of the console, so as to adjust the shooting angle of the high-definition camera, so that the image of the hoisting operation site captured by the high-definition camera contains the target to be hoisted, and the tracking target is always located at the center of each frame image of the continuous video frame when the real-time tracking of the tracking target is performed.

[0049] The high-definition camera is used to take a hoisting operation site image before the hoisting operation starts, and take continuous video frames of the tracking target after the tracking target is determined, so as to realize real-time tracking of the tracking target; wherein the hoisting operation site image at least contains one to-be-lifted target (i.e. the target that needs to be hoisted), and the moving target in the continuous video frames is the tracking target, which is the hoisting target that is currently being hoisted. The high-definition camera adjusts its parameters under the control of the console, and the parameters include the working mode and the focal length. When the hoisting operation is carried out in a low-visibility working condition (such as night, rain, fog, snow, etc.), the night vision mode or the infrared mode of the high-definition camera is started, so as to capture hoisting operation site images and continuous video frames with higher quality. The focal length is adjusted to improve the definition of each image in the hoisting operation site images and the continuous video frames, thereby improving the environmental adaptability of the auxiliary hoisting system.

[0050] The positioning device can be a GPS positioning module, and each to-be-lifted target and the just-in-place target corresponding to the to-be-lifted target is provided with a GPS positioning module, so as to realize spatial position positioning of the to-be-lifted target and the just-in-place target corresponding thereto; the positioning device can also be an ultra-wideband (UWB) real-time positioning network. Compared with the UWB real-time positioning network, the GPS positioning mode has the defects of high cost and low positioning accuracy, so the UWB real-time positioning network is adopted to realize real-time positioning of the hoisting target and the just-in-place target.

[0051] In one specific embodiment of the present application, the UWB real-time positioning network includes a UWB control module, M UWB base stations, and a plurality of UWB tags, M≥3; the M UWB base stations are connected with the UWB control module, and the UWB control module is connected with the console. In this embodiment, M=4, i.e. UWB base station A, UWB base station B, UWB base station C and UWB base station D.

[0052] The fan base position is the center of the hoisting operation area, and the UWB base station A, the UWB base station B, the UWB base station C and the UWB base station D are all arranged on the periphery of the hoisting operation area, so as to avoid forming an obstacle during the hoisting operation. The UWB base station A, the UWB base station B, the UWB base station C and the UWB base station D are located at the same height and are not collinear, and the distance between adjacent UWB base stations is 20-80 meters, so as to ensure the positioning accuracy. Specifically, as shown in the figure, Figure 1 and 2As shown, a three-dimensional self-defined coordinate system is constructed with UWB base station A as the origin, and UWB base station A, UWB base station B, UWB base station C and UWB base station D are located at the four vertices of the rectangle, which significantly improves the positioning accuracy; wherein UWB base station B is arranged on the x-axis of the self-defined coordinate system, UWB base station C is arranged on the y-axis of the self-defined coordinate system, UWB base station D is located on the xy plane of the self-defined coordinate system, and the z-axis is perpendicular to the xy plane, which can simplify the calculation of the spatial position. The spatial positions (coordinates) of UWB base station A, UWB base station B, UWB base station C and UWB base station D in the self-defined coordinate system can be determined, and the spatial positions of UWB base station A, UWB base station B, UWB base station C and UWB base station D in the WGS coordinate system can also be determined.

[0053] A UWB tag is arranged on each target to be hoisted and its corresponding target in place. Specifically, in order to improve the positioning accuracy, the UWB tag on the target to be hoisted is arranged on the mating surface of the target to be hoisted, and the UWB tag on the target in place is arranged on the mating surface of the target in place. The mating surface refers to the end surface that needs to be matched when the target to be hoisted is connected with the target in place. For example, during hoisting operation, the bottom surface of the target to be hoisted is connected with the top surface of the target in place, and the mating surface of the target to be hoisted refers to its bottom surface, and the mating surface of the target in place refers to its top surface.

[0054] Each UWB tag has a unique ID, and the ID of the UWB tag on the target to be hoisted corresponds to the ID of the UWB tag on the target in place matched with the target to be hoisted, so as to facilitate obtaining the spatial positions of the target to be hoisted (or the tracking target) and its corresponding target in place. In this embodiment, the ID of the UWB tag on the target to be hoisted is defined as Tn, and the ID of the UWB tag on the target in place corresponding to the target to be hoisted is defined as Tn+1, and n is an odd number. For example, the target to be hoisted is a tower drum, a nacelle or a hub, the ID of the UWB tag on the tower drum is defined as T1, the ID of the UWB tag on the target in place corresponding to the tower drum is defined as T2, the ID of the UWB tag on the nacelle is defined as T3, the ID of the UWB tag on the target in place corresponding to the nacelle is defined as T4, and the ID of the UWB tag on the hub is defined as T5, and the ID of the UWB tag on the target in place corresponding to the hub is defined as T6. When the tracking target is the tower drum, the communication time difference between the UWB tags with IDs of T1 and T2 and each UWB base station is obtained, and then the spatial positions of the tower drum and its corresponding target in place are calculated.

[0055] The UWB control module is configured to calculate distances between the tracking target and each UWB base station according to time differences in communication between a UWB tag on the tracking target and each UWB base station, and calculate distances between the in-place target corresponding to the tracking target and each UWB base station according to time differences in communication between a UWB tag on the in-place target and each UWB base station; obtain spatial positions of each UWB base station, and send the calculated distances and the spatial positions of each UWB base station to the console, so that the console calculates spatial positions of the tracking target and the in-place target corresponding to the tracking target according to the distances and the spatial positions of each UWB base station.

[0056] The spatial positions of the tracking target and the in-place target corresponding to the tracking target are obtained by using the UWB real-time positioning network, and the positioning accuracy can reach centimeter level after filtering the spatial positions, which greatly improves the hoisting and positioning accuracy and hoisting efficiency.

[0057] In the embodiment, the hoisting tool is a crane, and the console is arranged in a cab of the crane, so that the driver can guide the hoisting operation according to the hoisting process images (the hoisting site images and the continuous video frames of the tracking target) displayed on the console.

[0058] The console is configured to obtain and visualize the hoisting site images and the continuous video frames of the tracking target in real time, adjust parameters (including working modes and focal lengths) and shooting angles of the gimbal high-definition camera, so that the hoisting site images and the continuous video frames captured by the high-definition camera are clear, the hoisting site images contain at least one target to be hoisted, and the tracking target is always located at the center of each image in the continuous video frames. Before the hoisting starts, when the hoisting site images captured by the high-definition camera do not contain the target to be hoisted, the console adjusts the parameters and the shooting angles of the gimbal high-definition camera, so that the hoisting site images captured by the high-definition camera are clear and contain the target to be hoisted; after the tracking target is determined, when the continuous video frames captured by the high-definition camera do not contain the tracking target or the tracking target is not at the center of the image, the console adjusts the parameters and the shooting angles of the gimbal high-definition camera, so that each image in the continuous video frames captured by the high-definition camera is clear and the tracking target is located at the center of the image, which facilitates the driver to observe the continuous video frames when operating the crane, thereby guiding the hoisting operation of the tracking target.

[0059] The console is also used to identify all detection targets in the hoisting operation site image, and determine a tracking target from the detection targets. The console identifies the detection targets in the hoisting operation site image in the following manner: a target identification model is constructed in the console, and the target identification model is trained, then the trained target identification model is used to identify all detection targets in the hoisting operation site image, and a tracking target is determined from all the detection targets. In this embodiment, the detection target directly below the main hook of the crane is taken as the tracking target, or any detection target identified is manually selected as the tracking target. The detection targets include a target to be hoisted and a non-hoisted target. Exemplarily, the detection targets include a tower, a nacelle, a hub, and a person, wherein the person is a non-hoisted target, and the tower, the nacelle, and the hub are targets to be hoisted.

[0060] The console is also used to track the tracking target in real time according to the continuous video frames of the tracking target. In this embodiment, the tracking target is tracked in real time by using the target identification model and a deepsort algorithm. The implementation process is as follows: each frame of image of the continuous video frames is input to the target identification model to obtain a rectangular frame of the detection target, the rectangular frame (the coordinate values of the four vertices) of the detection target is taken as the input of the deepsort algorithm, the deepsort algorithm determines the same tracking target according to the rectangular frame of the detection target and the moving track of the detection target in the front and rear frames of images, and adjusts the shooting angle and parameters of the pan-tilt camera, thereby achieving real-time tracking of the tracking target.

[0061] The console is also used to obtain the spatial positions of the tracking target and the corresponding just-in-place target, calculate the real-time relative position information between the tracking target and the corresponding just-in-place target according to the spatial positions of the tracking target and the corresponding just-in-place target, and display the real-time relative position information and the corresponding frame of image in the continuous video frames on the display window of the console, so as to guide the hoisting operation of the tracking target.

[0062] In this embodiment, the real-time relative position information includes the absolute distance, the planar distance, and the height difference between the tracking target and the corresponding just-in-place target. When the corresponding frame of image of the tracking target is displayed, the specific numerical value of the real-time relative position information is displayed, as shown in FIG. 6. Figure 3 As shown in FIG. 6, the image of the tracking target and the corresponding relative position information are displayed on the display window of the console. The driver guides the hoisting operation of the tracking target according to the image and the information on the display window, thereby improving the hoisting efficiency and the just-in-place accuracy.

[0063] As shown in FIG. 6, the image of the tracking target and the corresponding relative position information are displayed on the display window of the console. The driver guides the hoisting operation of the tracking target according to the image and the information on the display window, thereby improving the hoisting efficiency and the just-in-place accuracy. Figure 4 As shown in FIG. 6, the image of the tracking target and the corresponding relative position information are displayed on the display window of the console. The driver guides the hoisting operation of the tracking target according to the image and the information on the display window, thereby improving the hoisting efficiency and the just-in-place accuracy.

[0064] Step 1: Obtain a hoisting operation site image, wherein the hoisting operation site image contains at least one target to be hoisted.

[0065] In this embodiment, a gimbal-mounted high-definition camera is used to acquire images of the hoisting operation site. When the images of the hoisting operation site are unclear and / or do not contain the target to be hoisted, the control console adjusts the parameters and shooting angle of the gimbal-mounted high-definition camera to make the captured images of the hoisting operation site clear and contain the target to be hoisted.

[0066] Step 2: Identify all detection targets in the hoisting operation site image and determine the tracking target from the detection targets.

[0067] Deploy the target recognition model in the console and use it to identify all detected targets in the hoisting operation site images. The specific implementation process includes:

[0068] Build an object recognition model and train it.

[0069] The trained target recognition model was used to identify all detected targets in the hoisting operation site images.

[0070] In this embodiment, the specific training process of the target recognition model is as follows:

[0071] (1) Constructing the sample dataset

[0072] A large number of wind turbine installation images can be obtained through two methods: first, installing high-definition cameras at the installation site to capture images of the wind turbine installation under various operating conditions (e.g., visible light, night vision, and infrared modes); second, crawling a large number of wind turbine installation images containing various models, shooting angles, and operating conditions from the internet. In this embodiment, approximately 1500 publicly available wind turbine installation images were collected from the internet using search engines such as Baidu and Sogou. These images cover various installation conditions and wind turbine types. Additionally, 500 images of key wind turbine components covering various installation stages were taken at the Yongzhou Xintian Jiufengshan Wind Farm using high-definition cameras. The high-definition camera used for image capture has two degrees of rotational freedom in both the horizontal and vertical planes and supports high-magnification focusing, enabling long-distance high-definition shooting.

[0073] The collected wind turbine hoisting images were manually annotated using the labelimg software. The annotations included the annotation category and pixel range. In this embodiment, the annotation categories included seven types: base, tower, hook, blade, nacelle, hub, and people (non-hoisting targets). Figure 5 The labelimg image shown is an annotation of key components of the wind turbine hoisting system. This image is a high-resolution camera image and includes three annotation categories or targets: base, tower, and person. These annotation categories or targets are the objects to be identified and detected later. The annotated images constitute the sample dataset.

[0074] (2) Construction and training of the target recognition model

[0075] According to the data amount of the labeled pictures, the target recognition accuracy and speed requirements, the YOLOV5 model is selected as the target recognition model; the pictures in the sample data set are taken as the model input, and the labeled categories and pixel ranges in the corresponding pictures are taken as the model output, the hyperparameter tuning is performed on the target recognition model, so as to obtain the trained target recognition model.

[0076] In the recognition or detection, the image of the hoisting operation site is captured by the gimbal high-definition camera, the input of the target recognition model is a color three-channel picture of 608*608, the model outputs all the prediction results of three scales (76*76, 38*38, 19*19), three output heads respectively predict three types of small, medium and large targets, and the dimension of 255 represents the category information and position information of the detection frame; then, the non-maximum suppression algorithm is used to eliminate most of the overlapping and low confidence detection frames, and the target recognition result is output.

[0077] Step 3: Obtain the continuous video frames of the tracking target, track the tracking target in real time according to the continuous video frames, and adjust the parameters and shooting angle of the gimbal high-definition camera, so that the tracking target is always located at the center of each frame of image.

[0078] The continuous video frames of the tracking target are obtained by using the gimbal high-definition camera, when a frame of image is not clear and / or the tracking target is not located at the center of the image, the parameters and shooting angle of the gimbal high-definition camera are adjusted by the control console, so that each frame of captured image is clear and the tracking target is located at the center of the image.

[0079] In this embodiment, the trained target recognition model and the deepsort algorithm are used to track the tracking target in real time, and the specific implementation process is as follows:

[0080] Each frame of image of the continuous video frames is input into the target recognition model to obtain the rectangular frame of the detection target;

[0081] The rectangular frame of the detection target is taken as the input of the deepsort algorithm, and the deepsort algorithm determines the tracking target in the corresponding frame of image according to the rectangular frame of the detection target and the moving track of the detection target in the front and rear frames of image;

[0082] The shooting angle and parameters of the gimbal high-definition camera are adjusted so that the tracking target is located at the center of the corresponding frame of image, and the tracking target is tracked in real time.

[0083] Step 4: Obtain the spatial positions of the tracking target and its corresponding just-in-place target, and calculate the real-time relative position information between the tracking target and its corresponding just-in-place target according to the spatial positions of the tracking target and its corresponding just-in-place target.

[0084] The UWB tag is arranged on each target to be hoisted and the corresponding target in position, the corresponding target in position is determined according to the ID of the UWB tag on the tracking target (i.e. the selected hoisting target), the distance between the target in position and each UWB base station can be calculated according to the communication time difference between the UWB tag on the target in position and each UWB base station, and then the spatial position of the target in position in the WGS coordinate system or the self-defined coordinate system can be calculated according to the spatial position of each UWB base station and the distance between the target in position and each UWB base station. Similarly, the spatial position of the tracking target in the WGS coordinate system or the self-defined coordinate system can be calculated.

[0085] In the embodiment, the relative position information includes the absolute distance, the planar distance and the height difference between the tracking target and the corresponding target in position, the absolute distance refers to the distance between the mating surface of the tracking target and the mating surface of the target in position, the planar distance refers to the distance between the tracking target and the target in position in the horizontal direction, and the height difference refers to the distance between the tracking target and the target in position in the vertical direction.

[0086] Step 5: display the real-time relative position information and the corresponding frame image in the continuous video frames to guide the hoisting operation of the tracking target.

[0087] The specific value of the real-time relative position information is displayed when the corresponding frame image of the tracking target is displayed, as shown in FIG. 6, when the hoisting operation of the tower drum is performed, the planar distance between the tower drum and the corresponding target in position is 40.3 m, the height difference is 21.9 m, and the absolute distance is 45.9 m. The image of the tracking target and the corresponding relative position information are displayed on the display window of the console, and the driver guides the hoisting operation of the tracking target according to the image and information on the display window, thereby improving the hoisting efficiency and the positioning accuracy. Figure 3

[0088] Step 6: repeat steps 1-5 to perform the hoisting operation of the next target to be hoisted until all the targets to be hoisted at the hoisting site are hoisted.

[0089] The above only discloses specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or modifications within the technical range disclosed by the present application, which shall be covered within the protection scope of the present application.​

Claims

1. A wind turbine generator system intelligent auxiliary hoisting system, characterized in that: The system comprises a gimbal high-definition camera, a positioning device, and a control console arranged in a hoisting tool; the gimbal high-definition camera and the positioning device are connected with the control console respectively; The gimbal high-definition camera is used for shooting hoisting site images and continuous video frames of a tracking target; The positioning device is used for positioning the spatial positions of the tracking target and its corresponding in-place target; The control console is used for acquiring and visualizing the hoisting site images and the continuous video frames of the tracking target in real time, adjusting the parameters and shooting angles of the gimbal high-definition camera, making the hoisting site images clear and containing at least one target to be hoisted, and making the continuous video frames contain the tracking target and the tracking target always located at the center of each frame image; identifying all detection targets in the hoisting site images, determining the tracking target, tracking the tracking target in real time according to the continuous video frames, acquiring the spatial positions of the tracking target and its corresponding in-place target, calculating real-time relative position information according to the spatial positions of the tracking target and its corresponding in-place target, and displaying the real-time relative position information and the corresponding frame images in the continuous video frames to guide the hoisting operation of the tracking target; The positioning device comprises an UWB control module, M UWB base stations, and a plurality of UWB tags, and M is greater than or equal to 3; the M UWB base stations are connected with the UWB control module respectively, and the UWB control module is connected with the control console; The M UWB base stations are arranged at the periphery of a hoisting operation area and are located at the same height and are not collinear; the UWB tags are arranged on the matching surfaces of each target to be hoisted and its corresponding in-place target, each UWB tag has a unique ID, and the ID of the UWB tag on the target to be hoisted corresponds to the ID of the UWB tag on the in-place target matched with the target to be hoisted; The UWB control module is used for calculating the distances between the tracking target and each UWB base station according to the communication time difference between the UWB tag on the tracking target and each UWB base station, and calculating the distances between the in-place target corresponding to the tracking target and each UWB base station according to the communication time difference between the UWB tag on the in-place target and each UWB base station; The spatial positions of each UWB base station are acquired, and the calculated distances and the spatial positions of each UWB base station are sent to the control console, so that the control console calculates the spatial positions of the tracking target and its corresponding in-place target according to the distances and the spatial positions of each UWB base station.

2. The wind turbine generator system intelligent auxiliary hoisting system according to claim 1, characterized in that: The spatial positions refer to the spatial positions in a WGS coordinate system or a self-defined coordinate system; The self-defined coordinate system is a three-dimensional coordinate system with a certain UWB base station as the origin and all UWB base stations located at the same height and not collinear.

3. A method of intelligent assisted hoisting of a wind turbine generator system, characterized in that, The method comprises the following steps: Acquiring hoisting site images, wherein the hoisting site images contain at least one target to be hoisted; Identifying all detection targets in the hoisting site images and determining a tracking target from the detection targets; Acquire the continuous video frames of the tracking target, track the tracking target in real time according to the continuous video frames, and adjust the parameters and shooting angle of the pan-tilt high-definition camera, so that the tracking target is always located at the center of each frame image; Acquire the spatial positions of the tracking target and its corresponding just-in-place target, and calculate the real-time relative position information between the tracking target and its corresponding just-in-place target according to the spatial positions of the tracking target and its corresponding just-in-place target; Display the real-time relative position information and the corresponding frame image in the continuous video frames to guide the hoisting operation of the tracking target; The spatial positions of the tracking target and its corresponding just-in-place target are positioned by a positioning device, the positioning device comprising a UWB control module, M UWB base stations and a plurality of UWB tags, M≥3; the M UWB base stations are respectively connected with the UWB control module, and the UWB control module is connected with a control console; The M UWB base stations are arranged at the periphery of the hoisting operation area and are located at the same height and are not collinear; the UWB tags are arranged on the matching surfaces of each target to be hoisted and its corresponding just-in-place target, each UWB tag has a unique ID, and the ID of the UWB tag on the target to be hoisted corresponds to the ID of the UWB tag on the just-in-place target matched with the target to be hoisted; The UWB control module is used to calculate the distance between the tracking target and each UWB base station according to the communication time difference between the UWB tag on the tracking target and each UWB base station, and calculate the distance between the just-in-place target corresponding to the tracking target and each UWB base station according to the communication time difference between the UWB tag on the just-in-place target and each UWB base station; Acquire the spatial positions of each UWB base station, and send the calculated distances and the spatial positions of each UWB base station to the control console, so that the control console calculates the spatial positions of the tracking target and its corresponding just-in-place target according to the distances and the spatial positions of each UWB base station.

4. The intelligent assisted hoisting method for a wind power generating set according to claim 3, characterized in that, The specific implementation process of identifying all detection targets in the hoisting operation site image is as follows: Construct a target identification model and train the target identification model; Identify all detection targets in the hoisting operation site image by using the trained target identification model.

5. The intelligent assisted hoisting method for a wind power generating set according to claim 4, characterized in that, The target identification model is a YOLOV5 model.

6. The smart assisted hoisting method of a wind power generator set according to claim 4 or 5, characterized in that, The target identification model and the deepsort algorithm are used to track the tracking target in real time, and the specific implementation process is as follows: Input each frame image of the continuous video frames into the target identification model to obtain the rectangular frame of the detection target; The rectangular frame of the detection target is input into the deepsort algorithm, and the deepsort algorithm determines the tracking target in the corresponding frame image according to the rectangular frame of the detection target and the moving track of the detection target in the front and rear frame images; Adjust the shooting angle and parameters of the pan-tilt high-definition camera, so that the tracking target is located at the center of the corresponding frame image.

7. The intelligent assisted lifting method of a wind power generator system according to claim 3, characterized in that, The detection target below the hoisting tool hook is taken as the tracking target, or any detection target is manually selected as the tracking target.

8. An electronic device, comprising: The electronic device comprises: a memory for storing a computer program; a processor for implementing the wind turbine generator unit intelligent auxiliary hoisting method according to any one of claims 3-7 when executing the computer program.

9. A computer-readable storage medium, characterized in that: The computer program is stored on the computer readable storage medium and is executed by the processor to implement the wind turbine generator unit intelligent auxiliary hoisting method according to any one of claims 3-7.

Images