A hoisting mechanism for a generator

Abstract

The present application relates to a kind of hoisting mechanism setting image acquisition module for collecting the image of generator in hoisting process for generating electricity, and image is sent to control host computer;Control host computer analyzes the motion trajectory of generator according to the image sent by image acquisition module in succession.The horizontal horizontal moving track data and vertical moving track data of control host computer are input into deep neural network model, and the working parameter of deep neural network model output is oscillating device, to offset the swing of generator;When image processing, the difference image of the front and rear two images is calculated, then the centroid position of net difference image is calculated, and the coordinates of centroid position are marked, and the coordinates are the coordinates of the pixel point where centroid is located;The continuous change of centroid position coordinates is saved as moving track data;Display module is set to display the working state of hoisting module and driving module, to ensure safety.

Description

Technical Field

[0001] This invention relates to the field of hoisting equipment, and more specifically to a hoisting mechanism for a generator. Background Technology

[0002] Large generators require hoisting equipment during assembly. However, current hoisting equipment sways during generator lifting, posing risks such as equipment instability or overturning due to excessive swaying, potentially causing injury or accidents. Collisions caused by swaying can damage the generator or other related equipment, impairing its quality and performance. Furthermore, the inability to effectively control and counteract swaying may necessitate greater time and effort to complete generator hoisting and assembly, leading to production delays.

[0003] Current solutions to this swaying problem generally involve updating or upgrading the lifting equipment, selecting equipment with better stability and control, such as cranes or robotic lifting equipment with automatic control systems. Introducing dampers or shock absorbers can reduce the swaying of the lifting equipment and increase stability. For example, installing dampers on the lifting ropes can reduce rope swaying. Using additional supports or fixing devices can also increase the stability of the lifting equipment. For example, temporary support columns can be installed or the lifting equipment can be secured to the ground to reduce its sway amplitude. Using technologies such as tilt sensors and vibration sensors to monitor the status and movement of the lifting equipment in real time allows for timely detection and handling of swaying problems.

[0004] However, the above solutions can only monitor and reduce sway, not eliminate it. Especially during assembly, the prolonged swaying can lengthen the assembly cycle and reduce production efficiency. Therefore, finding a way to quickly counteract the swaying of hoisting equipment is a pressing technical problem that needs to be solved. Summary of the Invention

[0005] To address the above problems, this invention provides a hoisting mechanism for a generator, comprising a control host, a hoisting module, a drive module, and an image acquisition module; characterized in that:

[0006] The control host connects to the hoisting module, drive module, and image acquisition module, and is used to control the operation of the hoisting module, drive module, and image acquisition module.

[0007] The hoisting module includes a drive motor and a reducer. The drive motor and reducer work together to drive the hoisting rope to rise and fall, thereby enabling the generator to be hoisted to rise and fall.

[0008] The drive module is used to drive the hoisting module to move along the horizontal X-axis and Y-axis, thereby enabling the generator to be hoisted to move horizontally;

[0009] The image acquisition module is used to acquire images of the generator during the hoisting process and send the images to the control host; the control host analyzes the generator's motion trajectory based on the images sent sequentially by the image acquisition module.

[0010] Furthermore, in one embodiment, it also includes an image analysis module and a display module, which are connected to the control host.

[0011] The image analysis module is used to process the images sent from the image acquisition module to the control host to obtain the processed images; the control host then analyzes the processed images to efficiently obtain the generator's motion trajectory.

[0012] The display module is used to show the working status of the lifting module and the drive module, and displays a prompt to stop working after the lifting module and the drive module have finished working.

[0013] Furthermore, in one embodiment, the hoisting mechanism further includes a traction module, which includes an X-axis traction device and a Y-axis traction device; the traction device is provided with a moving mechanism and a counterweight, and the position of the counterweight is changed by the moving mechanism to adjust the position of the center of mass in that direction; the traction module is installed under the hoisting module and moves together with the generator to be hoisted.

[0014] The X-axis vibratory puller includes an X-axis moving mechanism and an X-axis counterweight. The X-axis moving mechanism drives the X-axis counterweight to move along the X-axis, changing the position of the X-axis center of mass of the vibratory puller module. The Y-axis vibratory puller includes a Y-axis moving mechanism and a Y-axis counterweight. The Y-axis moving mechanism drives the Y-axis counterweight to move along the Y-axis, changing the position of the Y-axis center of mass of the vibratory puller module.

[0015] Furthermore, in one embodiment, the hoisting mechanism further includes a horizontal truss, on which two parallel horizontal guide rails are provided, and a connecting rail is provided between the horizontal guide rails and is arranged horizontally and perpendicularly to the horizontal guide rails; the drive module drives the connecting rail to move along the horizontal guide rails in the X-axis direction, and the drive module drives the hoisting module to move along the connecting rails in the Y-axis direction; thereby realizing that the drive module is used to drive the hoisting module to move along the horizontal X-axis direction and the Y-axis direction.

[0016] The drive motor of the hoisting module is equipped with a drive gear, which is connected to a reducer. The reducer is equipped with a drum, and the hoisting rope is wound around the drum. The lower part of the hoisting rope is connected to a clamp, which is connected to the generator to be hoisted. The rotation of the reducer drives the drum to rotate, and the generator to be hoisted is raised and lowered through the hoisting rope.

[0017] The traction module is installed on the gripper.

[0018] Furthermore, in one embodiment, the image acquisition module includes a horizontal camera and a vertical camera. The horizontal camera is arranged parallel to the gripper and is used to acquire images of the generator to be hoisted in the horizontal direction. The vertical camera is arranged below the hoisting module and is used to acquire a top view image of the top of the generator to be hoisted.

[0019] The horizontal and vertical cameras each capture 24-36 images per second. The image acquisition module sends the captured images to the control host, which then sends them to the image analysis module for image processing. The specific image processing process is as follows:

[0020] First, all images captured by the horizontal camera within 5-10 seconds are sequentially subtracted, i.e., the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of horizontal difference images. The horizontal difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net horizontal difference image that only includes the generator part. Next, the centroid position of the net horizontal difference image is calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as horizontal movement trajectory data.

[0021] Then, all images captured by the vertical camera within 5-10 seconds are successively subtracted, that is, the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of vertical difference images. The vertical difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net vertical difference image containing only the generator part. The centroid position of the net vertical difference image is then calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as vertical movement trajectory data.

[0022] The image analysis module sends the horizontal and vertical movement trajectory data to the control host.

[0023] Furthermore, in one embodiment, the control host inputs horizontal and vertical movement trajectory data into a deep neural network model, and the deep neural network model outputs the operating parameters of the X-axis and Y-axis vibrating devices.

[0024] The control host sends the operating parameters of the X-axis and Y-axis vibrating devices to the vibration module. The vibration module controls the operation of the X-axis and Y-axis vibrating devices according to the operating parameters, adjusting the center of mass of the X-axis and Y-axis to counteract the swaying of the generator.

[0025] Furthermore, in one implementation, the deep neural network model is a convolutional neural network model or a recurrent neural network model.

[0026] Furthermore, in one implementation, the modeling method for the deep neural network model is as follows:

[0027] After hoisting the generator using the lifting module, drive the generator to produce different forms of swaying and detect the swaying period; adjust the operation of the X-axis and Y-axis vibrating devices so that they drive the center of mass to reciprocate periodically, and ensure that the reciprocating period of the X-axis and Y-axis vibrating devices is the same as the swaying period of the generator; acquire and save images of the generator and the operating parameters of the X-axis and Y-axis vibrating devices.

[0028] Hoist generators of different weights and drive them to swing in different directions and angles. During each swing, change the operating parameters of the X-axis and Y-axis vibrating devices and measure the time required for the generator to stop swinging. Save the images of the generators that stop the fastest and the operating parameters of the X-axis and Y-axis vibrating devices as training samples.

[0029] The collected training samples are divided into a training set and a validation set. The deep neural network model is trained using the training set and then validated using the validation set. Once the validation is successful, the required deep neural network model is obtained.

[0030] A method for hoisting a generator, using the aforementioned generator hoisting mechanism, includes the following steps:

[0031] Step A: Train the deep neural network model:

[0032] After hoisting the generator using the lifting module, drive the generator to produce different forms of swaying and detect the swaying period; adjust the operation of the X-axis and Y-axis vibrating devices so that they drive the center of mass to reciprocate periodically, and ensure that the reciprocating period of the X-axis and Y-axis vibrating devices is the same as the swaying period of the generator; acquire and save images of the generator and the operating parameters of the X-axis and Y-axis vibrating devices.

[0033] Hoist generators of different weights and drive them to swing in different directions and angles. During each swing, change the operating parameters of the X-axis and Y-axis vibrating devices and measure the time required for the generator to stop swinging. Save the images of the generators that stop the fastest and the operating parameters of the X-axis and Y-axis vibrating devices as training samples.

[0034] The collected training samples are divided into a training set and a validation set. The deep neural network model is trained using the training set and then validated using the validation set. Once the validation is successful, the required deep neural network model is obtained.

[0035] Step B: Perform the actual hoisting:

[0036] The drive module drives the connecting rail to move along the horizontal guide rail in the X-axis direction, and the drive module drives the lifting module to move along the connecting rail in the Y-axis direction; thus, the drive module is used to drive the lifting module to move along the horizontal X-axis and Y-axis directions.

[0037] The drive motor of the hoisting module is equipped with a drive gear, which is connected to a reducer. The reducer is equipped with a drum, and the hoisting rope is wound around the drum. The lower part of the hoisting rope is connected to a clamp, which is connected to the generator to be hoisted. The rotation of the reducer drives the drum to rotate, and the generator to be hoisted is raised and lowered through the hoisting rope.

[0038] Step C: Acquire images:

[0039] The horizontal and vertical cameras each capture 24-36 images per second. The image acquisition module sends the captured images to the control host, which then sends them to the image analysis module for image processing. The specific image processing process is as follows:

[0040] First, all images captured by the horizontal camera within 5-10 seconds are sequentially subtracted, i.e., the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of horizontal difference images. The horizontal difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net horizontal difference image that only includes the generator part. Next, the centroid position of the net horizontal difference image is calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as horizontal movement trajectory data.

[0041] Then, all images captured by the vertical camera within 5-10 seconds are successively subtracted, that is, the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of vertical difference images. The vertical difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net vertical difference image containing only the generator part. The centroid position of the net vertical difference image is then calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as vertical movement trajectory data.

[0042] The image analysis module sends the horizontal and vertical movement trajectory data to the control host.

[0043] Step D, Counteracting the oscillation:

[0044] The control host inputs the horizontal and vertical movement trajectory data into the deep neural network model, and the deep neural network model outputs the working parameters of the X-axis and Y-axis vibrating devices.

[0045] The control host sends the operating parameters of the X-axis and Y-axis vibrating devices to the vibrating module. The vibrating module controls the operation of the X-axis and Y-axis vibrating devices according to the operating parameters, and adjusts the center of mass of the X-axis and Y-axis to counteract the swaying of the generator.

[0046] Step E, Display Status:

[0047] The display module shows the working status of the hoisting module and the drive module, and displays a stop prompt after the hoisting module, drive module and traction module have finished working to ensure safety.

[0048] Furthermore, in one implementation, the deep neural network model is a convolutional neural network model or a recurrent neural network model.

[0049] The beneficial effects of this invention are as follows:

[0050] This invention includes an image acquisition module for capturing images of the generator during hoisting and sending these images to the control host. The control host analyzes the generator's motion trajectory based on the images sequentially sent by the image acquisition module. The control host inputs horizontal and vertical movement trajectory data into a deep neural network model, which outputs operating parameters for the X-axis and Y-axis vibrating devices. The control host then sends these operating parameters to the vibrating module, which controls the X-axis and Y-axis vibrating devices based on these parameters, adjusting their centers of mass to counteract generator sway.

[0051] During image processing, this invention performs sequential subtraction on all images captured by the camera within 5-10 seconds, that is, by deleting the previous image from the next image, calculating the difference between the two images, and thus removing the background, thereby obtaining a series of difference images. The difference images are then filtered to completely remove the noise from the background portion of the difference, resulting in a net difference image that only includes the generator portion. The centroid position of the net difference image is then calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels at the centroid location. The continuous changes in the centroid position coordinates are saved as motion trajectory data.

[0052] The present invention also includes a display module to show the working status of the hoisting module and the drive module, and the display module displays a stop prompt after the hoisting module, drive module and traction module have finished working to ensure safety. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Appendix Figure 1 This is a schematic diagram of the overall architecture of the present invention;

[0055] Appendix Figure 2 This is a structural diagram of the hoisting mechanism of the present invention;

[0056] Appendix Figure 3 This is a structural diagram of the hoisting module of the present invention. Detailed Implementation

[0057] Example 1:

[0058] See Figure 1-3 This invention provides a hoisting mechanism for a generator, comprising a control host, a hoisting module 3, a drive module, and an image acquisition module; characterized in that:

[0059] The control host connects to the hoisting module 3, the drive module, and the image acquisition module, and is used to control the operation of the hoisting module 3, the drive module, and the image acquisition module;

[0060] The hoisting module 3 includes a drive motor 4 and a reducer 5. The drive motor 4 and the reducer 5 work together to drive the hoisting rope 6 to rise and fall, thereby enabling the generator to be hoisted to rise and fall.

[0061] The drive module is used to drive the hoisting module 3 to move along the horizontal X-axis and Y-axis, thereby enabling the generator to be hoisted to move horizontally;

[0062] The image acquisition module is used to acquire images of the generator during the hoisting process and send the images to the control host; the control host analyzes the generator's motion trajectory based on the images sent sequentially by the image acquisition module.

[0063] Furthermore, in one embodiment, it also includes an image analysis module and a display module, which are connected to the control host.

[0064] The image analysis module is used to process the images sent from the image acquisition module to the control host to obtain the processed images; the control host then analyzes the processed images to efficiently obtain the generator's motion trajectory.

[0065] The display module is used to show the working status of the lifting module 3 and the drive module, and the display module displays a prompt to stop working after the lifting module 3 and the drive module have finished working.

[0066] Furthermore, in one embodiment, the hoisting mechanism further includes a traction module, which includes an X-axis traction device and a Y-axis traction device; the traction device is provided with a moving mechanism and a counterweight, and the position of the counterweight is changed by the moving mechanism to adjust the position of the center of mass in that direction; the traction module is installed under the hoisting module 3 and moves together with the generator to be hoisted.

[0067] The X-axis vibratory puller includes an X-axis moving mechanism and an X-axis counterweight. The X-axis moving mechanism drives the X-axis counterweight to move along the X-axis, changing the position of the X-axis center of mass of the vibratory puller module. The Y-axis vibratory puller includes a Y-axis moving mechanism and a Y-axis counterweight. The Y-axis moving mechanism drives the Y-axis counterweight to move along the Y-axis, changing the position of the Y-axis center of mass of the vibratory puller module.

[0068] Furthermore, in one embodiment, the hoisting mechanism further includes a horizontal truss, on which two parallel horizontal guide rails 1 are arranged, and a connecting rail 2 arranged horizontally and perpendicularly to the horizontal guide rails 1 is arranged between the horizontal guide rails 1; the drive module drives the connecting rail 2 to move along the horizontal guide rails 1 in the X-axis direction, and the drive module drives the hoisting module 3 to move along the connecting rail 2 in the Y-axis direction; thereby realizing that the drive module is used to drive the hoisting module 3 to move along the horizontal X-axis direction and the Y-axis direction.

[0069] The drive motor 4 of the hoisting module 3 is equipped with a drive gear, which is connected to the reducer 5. The reducer 5 is equipped with a drum, and the hoisting rope 6 is wound around the drum. The lower part of the hoisting rope 6 is connected to the gripper, which is connected to the generator to be hoisted. The rotation of the reducer 5 drives the drum to rotate, and the hoisting rope 6 drives the generator to be hoisted to rise and fall.

[0070] The traction module is installed on the gripper.

[0071] Furthermore, in one embodiment, the image acquisition module includes a horizontal camera and a vertical camera. The horizontal camera is set parallel to the gripper and is used to acquire images of the generator to be hoisted in the horizontal direction. The vertical camera is set below the hoisting module 3 and is used to acquire a top view image of the top of the generator to be hoisted.

[0072] The horizontal and vertical cameras each capture 24-36 images per second. The image acquisition module sends the captured images to the control host, which then sends them to the image analysis module for image processing. The specific image processing process is as follows:

[0073] First, all images captured by the horizontal camera within 5-10 seconds are sequentially subtracted, i.e., the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of horizontal difference images. The horizontal difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net horizontal difference image that only includes the generator part. Next, the centroid position of the net horizontal difference image is calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as horizontal movement trajectory data.

[0074] Then, all images captured by the vertical camera within 5-10 seconds are successively subtracted, that is, the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of vertical difference images. The vertical difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net vertical difference image containing only the generator part. The centroid position of the net vertical difference image is then calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as vertical movement trajectory data.

[0075] The image analysis module sends the horizontal and vertical movement trajectory data to the control host.

[0076] Furthermore, in one embodiment, the control host inputs horizontal and vertical movement trajectory data into a deep neural network model, and the deep neural network model outputs the operating parameters of the X-axis and Y-axis vibrating devices.

[0077] The control host sends the operating parameters of the X-axis and Y-axis vibrating devices to the vibration module. The vibration module controls the operation of the X-axis and Y-axis vibrating devices according to the operating parameters, adjusting the center of mass of the X-axis and Y-axis to counteract the swaying of the generator.

[0078] Furthermore, in one implementation, the deep neural network model is a convolutional neural network model or a recurrent neural network model.

[0079] Furthermore, in one implementation, the modeling method for the deep neural network model is as follows:

[0080] After hoisting the generator using lifting module 3, drive the generator to produce different forms of swaying and detect the swaying period; adjust the operation of the X-axis and Y-axis vibrating devices so that the X-axis and Y-axis vibrating devices drive the center of mass to reciprocate periodically, and ensure that the reciprocating period of the X-axis and Y-axis vibrating devices is the same as the swaying period of the generator; acquire and save images of the generator and the operating parameters of the X-axis and Y-axis vibrating devices.

[0081] Hoist generators of different weights and drive them to swing in different directions and angles. During each swing, change the operating parameters of the X-axis and Y-axis vibrating devices and measure the time required for the generator to stop swinging. Save the images of the generators that stop the fastest and the operating parameters of the X-axis and Y-axis vibrating devices as training samples.

[0082] The collected training samples are divided into a training set and a validation set. The deep neural network model is trained using the training set and then validated using the validation set. Once the validation is successful, the required deep neural network model is obtained.

[0083] Example 2:

[0084] A method for hoisting a generator, using the aforementioned generator hoisting mechanism, includes the following steps:

[0085] Step A: Train the deep neural network model:

[0086] After hoisting the generator using lifting module 3, drive the generator to produce different forms of swaying and detect the swaying period; adjust the operation of the X-axis and Y-axis vibrating devices so that the X-axis and Y-axis vibrating devices drive the center of mass to reciprocate periodically, and ensure that the reciprocating period of the X-axis and Y-axis vibrating devices is the same as the swaying period of the generator; acquire and save images of the generator and the operating parameters of the X-axis and Y-axis vibrating devices.

[0087] Hoist generators of different weights and drive them to swing in different directions and angles. During each swing, change the operating parameters of the X-axis and Y-axis vibrating devices and measure the time required for the generator to stop swinging. Save the images of the generators that stop the fastest and the operating parameters of the X-axis and Y-axis vibrating devices as training samples.

[0088] The collected training samples are divided into a training set and a validation set. The deep neural network model is trained using the training set and then validated using the validation set. Once the validation is successful, the required deep neural network model is obtained.

[0089] Step B: Perform the actual hoisting:

[0090] The drive module drives the connecting rail 2 to move along the horizontal guide rail 1 in the X-axis direction, and the drive module drives the lifting module 3 to move along the connecting rail 2 in the Y-axis direction; thus, the drive module is used to drive the lifting module 3 to move along the horizontal X-axis and Y-axis directions.

[0091] The drive motor 4 of the hoisting module 3 is equipped with a drive gear, which is connected to the reducer 5. The reducer 5 is equipped with a drum, and the hoisting rope 6 is wound around the drum. The lower part of the hoisting rope 6 is connected to the gripper, which is connected to the generator to be hoisted. The rotation of the reducer 5 drives the drum to rotate, and the hoisting rope 6 drives the generator to be hoisted to rise and fall.

[0092] Step C: Acquire images:

[0093] The horizontal and vertical cameras each capture 24-36 images per second. The image acquisition module sends the captured images to the control host, which then sends them to the image analysis module for image processing. The specific image processing process is as follows:

[0094] First, all images captured by the horizontal camera within 5-10 seconds are sequentially subtracted, i.e., the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of horizontal difference images. The horizontal difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net horizontal difference image that only includes the generator part. Next, the centroid position of the net horizontal difference image is calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as horizontal movement trajectory data.

[0095] Then, all images captured by the vertical camera within 5-10 seconds are successively subtracted, that is, the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of vertical difference images. The vertical difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net vertical difference image containing only the generator part. The centroid position of the net vertical difference image is then calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as vertical movement trajectory data.

[0096] The image analysis module sends the horizontal and vertical movement trajectory data to the control host.

[0097] Step D, Counteracting the oscillation:

[0098] The control host inputs the horizontal and vertical movement trajectory data into the deep neural network model, and the deep neural network model outputs the working parameters of the X-axis and Y-axis vibrating devices.

[0099] The control host sends the operating parameters of the X-axis and Y-axis vibrating devices to the vibrating module. The vibrating module controls the operation of the X-axis and Y-axis vibrating devices according to the operating parameters, and adjusts the center of mass of the X-axis and Y-axis to counteract the swaying of the generator.

[0100] Step E, Display Status:

[0101] The display module shows the working status of the hoisting module 3 and the drive module. After the hoisting module 3, the drive module and the traction module have finished working, the display module will show a prompt to stop working to ensure safety.

[0102] Furthermore, in one implementation, the deep neural network model is a convolutional neural network model or a recurrent neural network model.

[0103] Thus far, the description of the above embodiments has been provided for illustrative and descriptive purposes. This is not intended to be exhaustive or limiting of the present disclosure. Individual elements or features of particular embodiments are generally not limited to those particular embodiments, but may be interchanged and used in selected embodiments where applicable, even if not specifically shown or described. In many respects, the same elements or features may also be varied. Such variations are not considered a departure from this disclosure, and all such modifications are intended to be included within the scope of this disclosure.

[0104] Example embodiments are provided so that this disclosure will become thorough and will fully convey the scope to those skilled in the art. Numerous details, such as examples of specific parts, apparatus, and methods, are set forth to provide a thorough understanding of embodiments of this disclosure. It will be apparent to those skilled in the art that the specific details are not required, and the example embodiments may be implemented in many different forms, neither of which should be construed as limiting the scope of this disclosure. In some example embodiments, well-known processes, well-known apparatus structures, and well-known techniques are not described in detail.

[0105] Technical terms are used herein for the purpose of describing specific exemplary embodiments only and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms “a” and “the” as used herein may also refer to the plural forms. The terms “comprising” and “having” are inclusive and therefore specify the presence of the stated features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or additional having of one or more other features, integrals, steps, operations, elements, components, and / or combinations thereof. Unless expressly indicated in order of execution, the method steps, processes, and operations described herein are not to be construed as necessarily requiring performance in the specific order discussed and shown. It should also be understood that additional or optional steps may be employed.

Claims

1. A hoisting mechanism for a generator, comprising a control host, a hoisting module (3), a drive module, and an image acquisition module; characterized in that: The control host is connected to the hoisting module (3), the drive module and the image acquisition module, and is used to control the operation of the hoisting module (3), the drive module and the image acquisition module; The hoisting module (3) includes a drive motor (4) and a reducer (5). The drive motor (4) and the reducer (5) work together to drive the hoisting rope (6) to rise and fall, thereby driving the generator to be hoisted to rise and fall. The drive module is used to drive the hoisting module (3) to move along the horizontal X-axis and Y-axis, thereby driving the generator to be hoisted to move horizontally; The image acquisition module is used to acquire images of the generator during the hoisting process and send the images to the control host; the control host analyzes the generator's movement trajectory based on the images sent sequentially by the image acquisition module. It also includes an image analysis module and a display module, which are connected to the control host. The image analysis module is used to process the images sent from the image acquisition module to the control host to obtain the processed images; the control host then analyzes the processed images to efficiently obtain the generator's motion trajectory. The display module is used to display the working status of the lifting module (3) and the drive module, and the display module displays a prompt to stop working after the lifting module (3) and the drive module have finished working; The hoisting mechanism further includes a vibrating drag module, which includes an X-axis vibrating dragger and a Y-axis vibrating dragger; the vibrating dragger is equipped with a moving mechanism and a counterweight, and the position of the counterweight is changed by the moving mechanism to adjust the position of the center of mass in that direction; the vibrating drag module is installed under the hoisting module (3) and moves together with the generator to be hoisted; The X-axis vibratory puller includes an X-axis moving mechanism and an X-axis counterweight. The X-axis moving mechanism drives the X-axis counterweight to move along the X-axis, changing the position of the X-axis center of mass of the vibratory puller module. The Y-axis vibratory puller includes a Y-axis moving mechanism and a Y-axis counterweight. The Y-axis moving mechanism drives the Y-axis counterweight to move along the Y-axis, changing the position of the Y-axis center of mass of the vibratory puller module.

2. The generator hoisting mechanism according to claim 1, characterized in that: The hoisting mechanism also includes a horizontal truss, on which two parallel horizontal guide rails (1) are set, and a connecting rail (2) is set between the horizontal guide rails (1) and is perpendicular to the horizontal guide rails (1); the drive module drives the connecting rail (2) to move along the horizontal guide rail (1) in the X-axis direction, and the drive module drives the hoisting module (3) to move along the connecting rail (2) in the Y-axis direction; thus, the drive module is used to drive the hoisting module (3) to move along the horizontal X-axis direction and the Y-axis direction. The drive motor (4) of the hoisting module (3) is equipped with a drive gear, which is connected to a reducer (5); the reducer (5) is equipped with a drum, the hoisting rope (6) is wound around the drum, the lower part of the hoisting rope (6) is connected to a gripper, the gripper is connected to the generator to be hoisted, the reducer (5) rotates to drive the drum to rotate, and the generator to be hoisted is lifted and lowered through the hoisting rope (6); The traction module is installed on the gripper.

3. The generator hoisting mechanism according to claim 1, characterized in that: The image acquisition module includes a horizontal camera and a vertical camera. The horizontal camera is set parallel to the gripper and is used to acquire images of the generator to be hoisted in the horizontal direction. The vertical camera is set below the hoisting module (3) and is used to acquire a top view image of the generator to be hoisted from the top. The horizontal and vertical cameras each capture 24-36 images per second. The image acquisition module sends the captured images to the control host, which then sends them to the image analysis module for image processing. The specific image processing process is as follows: First, all images captured by the horizontal camera within 5-10 seconds are sequentially subtracted, i.e., the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of horizontal difference images. The horizontal difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net horizontal difference image that only includes the generator part. Next, the centroid position of the net horizontal difference image is calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as horizontal movement trajectory data. Then, all images captured by the vertical camera within 5-10 seconds are successively subtracted, that is, the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of vertical difference images. The vertical difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net vertical difference image containing only the generator part. The centroid position of the net vertical difference image is then calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as vertical movement trajectory data. The image analysis module sends the horizontal and vertical movement trajectory data to the control host.

4. The generator hoisting mechanism according to claim 3, characterized in that: The control host inputs the horizontal and vertical movement trajectory data into the deep neural network model, and the deep neural network model outputs the operating parameters of the X-axis and Y-axis vibrating devices. The control host sends the operating parameters of the X-axis and Y-axis vibrating devices to the vibration module. The vibration module controls the operation of the X-axis and Y-axis vibrating devices according to the operating parameters, adjusting the center of mass of the X-axis and Y-axis to counteract the swaying of the generator.

5. The generator hoisting mechanism according to claim 4, characterized in that: Deep neural network models are either convolutional neural network models or recurrent neural network models.

6. The generator hoisting mechanism according to claim 4, characterized in that: The modeling method for deep neural network models is as follows: After the generator is hoisted using the hoisting module (3), it is driven to produce different forms of swaying, and the swaying period of the generator is detected; the operation of the X-axis vibrating device and the Y-axis vibrating device is adjusted so that the X-axis vibrating device and the Y-axis vibrating device drive the center of mass to reciprocate periodically, and the reciprocating period of the X-axis vibrating device and the Y-axis vibrating device is the same as the swaying period of the generator; the image of the generator and the working parameters of the X-axis vibrating device and the Y-axis vibrating device are collected and saved; Hoist generators of different weights and drive them to swing in different directions and angles. During each swing, switch the operating parameters of the X-axis and Y-axis vibrators and measure the time required for the generator to stop swinging. Save the images of the generators that stop the fastest, along with the operating parameters of the X-axis and Y-axis vibrating oscillators, as training samples. The collected training samples are divided into a training set and a validation set. The deep neural network model is trained using the training set and then validated using the validation set. Once the validation is successful, the required deep neural network model is obtained.

7. A method for hoisting a generator, using the generator hoisting mechanism as described in claim 6, characterized in that... Includes the following steps: Step A: Train the deep neural network model: After the generator is hoisted using the hoisting module (3), it is driven to produce different forms of swaying, and the swaying period of the generator is detected; the operation of the X-axis vibrating device and the Y-axis vibrating device is adjusted so that the X-axis vibrating device and the Y-axis vibrating device drive the center of mass to reciprocate periodically, and the reciprocating period of the X-axis vibrating device and the Y-axis vibrating device is the same as the swaying period of the generator; the image of the generator and the working parameters of the X-axis vibrating device and the Y-axis vibrating device are collected and saved; Hoist generators of different weights and drive them to swing in different directions and angles. During each swing, switch the operating parameters of the X-axis and Y-axis vibrators and measure the time required for the generator to stop swinging. Save the images of the generators that stop the fastest, along with the operating parameters of the X-axis and Y-axis vibrating oscillators, as training samples. The collected training samples are divided into a training set and a validation set. The deep neural network model is trained using the training set and then validated using the validation set. Once the validation is successful, the required deep neural network model is obtained. Step B: Perform the actual hoisting: The drive module drives the connecting rail (2) to move along the horizontal guide rail (1) in the X-axis direction, and the drive module drives the lifting module (3) to move along the connecting rail (2) in the Y-axis direction; thus, the drive module is used to drive the lifting module (3) to move along the horizontal X-axis and Y-axis directions. The drive motor (4) of the hoisting module (3) is equipped with a drive gear, which is connected to a reducer (5); the reducer (5) is equipped with a drum, the hoisting rope (6) is wound around the drum, the lower part of the hoisting rope (6) is connected to a gripper, the gripper is connected to the generator to be hoisted, the reducer (5) rotates to drive the drum to rotate, and the generator to be hoisted is lifted and lowered through the hoisting rope (6); Step C: Acquire images: The horizontal and vertical cameras each capture 24-36 images per second. The image acquisition module sends the captured images to the control host, which then sends them to the image analysis module for image processing. The specific image processing process is as follows: First, all images captured by the horizontal camera within 5-10 seconds are sequentially subtracted, i.e., the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of horizontal difference images. The horizontal difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net horizontal difference image that only includes the generator part. Next, the centroid position of the net horizontal difference image is calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as horizontal movement trajectory data. Then, all images captured by the vertical camera within 5-10 seconds are successively subtracted, that is, the previous image is deleted from the next image, and the difference between the two images is calculated to remove the background, thus obtaining a series of vertical difference images. The vertical difference images are then filtered to completely remove the noise from the background subtraction, resulting in a net vertical difference image containing only the generator part. The centroid position of the net vertical difference image is then calculated, and the coordinates of the centroid position are marked, which are the coordinates of the pixels where the centroid is located. The continuous change of the centroid position coordinates is saved as vertical movement trajectory data. The image analysis module sends the horizontal and vertical movement trajectory data to the control host. Step D, Counteracting the oscillation: The control host inputs the horizontal and vertical movement trajectory data into the deep neural network model, and the deep neural network model outputs the operating parameters of the X-axis and Y-axis vibrating devices. The control host sends the operating parameters of the X-axis and Y-axis vibrating devices to the vibrating module. The vibrating module controls the operation of the X-axis and Y-axis vibrating devices according to the operating parameters, and adjusts the center of mass of the X-axis and Y-axis to counteract the swaying of the generator. Step E, Display Status: The display module shows the working status of the hoisting module (3) and the drive module. After the hoisting module (3), the drive module and the traction module have finished working, the display module will display a prompt to stop working to ensure safety.

8. The hoisting method for a generator as described in claim 7, characterized in that: Deep neural network models are either convolutional neural network models or recurrent neural network models.

Images