Closed-loop anti-sway system and method for grab-type ship unloader, and grab-type ship unloader

Abstract

The present invention proposes a closed-loop anti-sway system, method and grab-type ship unloader of a grab-type ship unloader, wherein the closed-loop anti-sway system of the grab-type ship unloader includes two grab cameras, a video processing server, an anti-sway controller and a crane main controller. The video processing server extracts the grab from the video information and generates the pixel coordinates of the grab. The video processing server calculates the actual spatial coordinates of the grab in combination with the pixel coordinates of the grab and the working condition data. The anti-sway controller adjusts the movement of the trolley and the rope according to the real-time additional speed so that the current position of the grab is controlled within a preset range. The closed-loop anti-sway system, method and grab-type ship unloader of the grab-type ship unloader of the present invention can effectively prevent the grab of the grab ship unloader from shaking during operation, and have the advantages of low production cost and low maintenance cost.

Description

Technical Field

[0001] The present invention relates to the field of mechanical engineering technology, and in particular to a closed-loop anti-sway system for a grab ship unloader, a closed-loop anti-sway method for a grab ship unloader, a grab ship unloader, an edge computing device, a computer-readable medium, a computer program, and a computer program product. Background Art

[0002] Grab ship unloader (GSU) can be used to take out cargo (e.g. coal) from the ship on the shore and place it in a transport device. The precise positioning and anti-sway control of the grab bucket of the grab ship unloader are key control links. During the operation of the grab ship unloader, the grab bucket is controlled only by the tension of the wire rope, and different degrees of swing will occur during the lifting process and the trolley travel process. If the control is improper, it will affect the operating efficiency at the least, and even cause collisions and other dangerous situations. Therefore, anti-sway control is required regardless of manual, semi-automatic or fully automatic operation.

[0003] In the traditional manual operation process, the ship unloader driver usually relies on the years of experience to accurately control the speed and lifting speed of the ship unloader and achieve manual anti-sway. Due to long-term and frequent operations and technical dependence on the operator, operating errors and dangerous accidents are inevitable. With the gradual development of science and technology, people try to upgrade the anti-sway system of grab ship unloaders, usually through the following aspects of technical improvement and innovation:

[0004] a) Open-loop anti-sway method: This method is the most common and is more practical when the grab does not swing at the starting point. However, it has low control accuracy and slow response. In addition, when there is a swing angle at the starting point, the open-loop anti-sway method is prone to failure. This method can only be combined with manual intervention and used under certain working conditions.

[0005] b) Installing a GPS inertial navigation system on the grab bucket, positioning the grab bucket through GPS technology, and realizing closed-loop anti-sway control. This method can greatly improve the anti-sway effect, but when the grab bucket is below the deck, the GPS signal is weak and the response is unstable. This method requires the internal modification of the grab bucket and the addition of battery power supply, which is difficult to implement and maintain later.

[0006] c) Add a light source or reflective strip above the grab, install a photosensitive receiver under the trolley frame, and calculate the swing amplitude of the grab by detecting the light source or reflected light, thereby realizing closed-loop control of the grab. In the process of executing this method, the trolley vibrates greatly, affecting the detection accuracy. The trolley power supply needs to be modified to add a trolley busbar or a towing device. The light source installed on the grab also needs to be powered by a battery. If a reflective strip is used, it is easy to be covered or rubbed off by the material during long-term operation. This method has high modification costs and high maintenance costs in the later stage. Summary of the invention

[0007] In view of this, the purpose of the present invention is to propose a closed-loop anti-sway system for a grab ship unloader, a closed-loop anti-sway method for a grab ship unloader, a grab ship unloader, an edge computing device, a computer readable medium, a computer program and a computer program product, which can effectively prevent the grab of the grab ship unloader from shaking during operation and have the advantages of low production cost and low maintenance cost.

[0008] The present invention proposes a closed-loop anti-sway system for a grab-type ship unloader, which is applied to the grab-type ship unloader. The grab-type ship unloader includes a trolley, a grab bucket and a main beam. The grab bucket is located below the trolley and is connected to the trolley through a rope. The trolley can travel and move on the main beam. The closed-loop anti-sway system for the grab-type ship unloader includes two grab bucket cameras, a video processing server, an anti-sway controller and a crane main controller. The two grab bucket cameras are arranged at both ends of the travel track of the trolley on the main beam. The field of view of the grab bucket camera covers the movement range of the grab bucket during the operation of the grab-type ship unloader. The grab bucket camera collects video information during the operation of the grab bucket unloader. The video processing server receives video information from the grab bucket camera, extracts the grab bucket from the video information, and generates pixel coordinates of the grab bucket. Wherein, the video processing server is connected to the anti-sway controller and the crane main controller, and the video processing server calculates the actual spatial coordinates of the grab bucket by combining the pixel coordinates of the grab bucket and the working condition data from the crane main controller; the anti-sway controller obtains the coordinate information and operation instruction information of the trolley, the length information and operation instruction information of the rope from the crane main controller, and calculates the real-time additional speed of the trolley and the real-time additional speed of the rope based on this information and the actual spatial coordinates of the grab bucket, and sends the real-time additional speed to the crane main controller, and the crane main controller adjusts the movement of the trolley and the rope according to the real-time additional speed so that the current position of the grab bucket is controlled within a preset range.

[0009] In an illustrative embodiment of a closed-loop anti-sway system for a grab ship unloader, the video processing server includes a preliminary processing module, a training module, and a detection module. The preliminary processing module receives video information from the grab camera, extracts different frames in different video information at intervals to obtain more than one thousand original frame images for training the algorithm, and performs target annotation on various postures of the grab by calibration to generate an annotation file. The training module organizes the annotation file into a deep learning training set, and then executes the training algorithm using the deep learning training set as the true value to obtain a training result weight file. The detection module receives real-time video information and uses the training result weight file to extract the grab from the video information and generate the pixel coordinates of the grab, and calculates the actual spatial coordinates of the grab by combining the pixel coordinates of the grab and the working condition data from the crane main controller.

[0010] The present invention also provides a grab-type ship unloader, which comprises the above-mentioned closed-loop anti-sway system of the grab-type ship unloader.

[0011] In an illustrative embodiment of a grab ship unloader, the video processing server includes a preliminary processing module, a training module and a detection module. The preliminary processing module receives video information from the grab camera, extracts different frames in different video information at intervals to obtain more than one thousand original frame images for training the algorithm, and marks various postures of the grab by calibration to generate a marking file. The training module organizes the marking file into a deep learning training set, and then executes the training algorithm using the deep learning training set as the true value to obtain a training result weight file. The detection module receives real-time video information and uses the training result weight file to extract the grab from the video information and generate the pixel coordinates of the grab, and calculates the actual spatial coordinates of the grab by combining the pixel coordinates of the grab and the working condition data from the crane main controller.

[0012] The present invention further proposes a closed-loop anti-sway method for a grab-type ship unloader, which is applied to the grab-type ship unloader. The grab-type ship unloader comprises a trolley, a grab bucket and a main beam. The grab bucket is located below the trolley and is connected to the trolley through a rope. The trolley can travel and move on the main beam. The closed-loop anti-sway method for the grab-type ship unloader comprises the following steps:

[0013] Collecting video information of the grab-type ship unloader during operation, wherein the field of view of the video information covers the movement range of the grab bucket of the grab-type ship unloader during operation;

[0014] Extracting the grab from the video information and generating pixel coordinates of the grab, and combining the pixel coordinates of the grab with the working condition data to calculate and obtain the actual spatial coordinates of the grab;

[0015] According to the coordinate information and operation instruction information of the trolley, the length information and operation instruction information of the rope and the actual spatial coordinates of the grab bucket, the real-time additional speed of the trolley and the real-time additional speed of the rope are calculated, and the movement of the trolley and the rope is adjusted according to the real-time additional speed so that the current position of the grab bucket is controlled within a preset range.

[0016] In an illustrative embodiment of a closed-loop anti-sway method for a grab-type ship unloader, the closed-loop anti-sway method for a grab-type ship unloader further comprises the following steps:

[0017] Receiving video information from the grab bucket camera, extracting different frames in different video information at intervals to obtain more than one thousand original frame images for training the algorithm, and marking various postures of the grab bucket by calibration to generate a marking file;

[0018] Arranging the annotated files into a deep learning training set, and then executing a training algorithm using the deep learning training set as a true value to obtain a training result weight file; and

[0019] Receive real-time video information, and use the training result weight file to extract the grab from the real-time video information and generate pixel coordinates of the grab.

[0020] The present invention also proposes an edge computing device, which has at least one memory and at least one processor; the at least one memory is used to store a machine-readable program; the at least one processor is used to call the machine-readable program to execute any one of the above-mentioned closed-loop anti-sway methods for a grab ship unloader.

[0021] The present invention further proposes a computer-readable medium, characterized in that computer instructions are stored on the computer-readable medium, and when the computer instructions are executed by a processor, the processor executes any one of the above-mentioned closed-loop anti-sway methods for a grab-type ship unloader.

[0022] The present invention also proposes a computer program, which includes computer executable instructions, and when the computer executable instructions are executed, at least one processor executes any one of the above-mentioned closed-loop anti-sway methods for a grab-type ship unloader.

[0023] The present invention further proposes a computer program product, which is tangibly stored on a computer-readable medium and includes computer-executable instructions, which, when executed, enable at least one processor to perform any of the above-mentioned closed-loop anti-sway methods for a grab-type ship unloader.

[0024] In the closed-loop anti-sway system, closed-loop anti-sway method, grab ship unloader, edge computing device, computer-readable medium, computer program and computer program product of the grab ship unloader of the present invention, video information of the grab ship unloader during operation is collected, the grab is extracted from the video information, and the pixel coordinates of the grab are generated. The actual space coordinates of the grab are calculated by combining the pixel coordinates of the grab and the working condition data. Then, the real-time additional speed of the trolley and the real-time additional speed of the rope are calculated according to the coordinate information and operation instruction information of the trolley, the length information and operation instruction information of the rope and the actual space coordinates of the grab. The movement of the trolley and the rope is adjusted according to the real-time additional speed so that the current position of the grab is controlled within a preset range, thereby effectively preventing the grab from shaking during operation. The present invention can achieve lower production costs and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings so that those skilled in the art can better understand the above and other features and advantages of the present invention. In the accompanying drawings:

[0026] Figure 1 It is a schematic diagram of the closed-loop anti-sway system of a grab ship unloader of the present invention when it is applied to the grab ship unloader.

[0027] Figure 2 It is a schematic diagram of the architecture of the closed-loop anti-sway system of the grab ship unloader of the present invention.

[0028] Figure 3 for Figure 2 The schematic diagram of the architecture of the video processing server of the closed-loop anti-sway system of the grab ship unloader is shown.

[0029] Figure 4 This is a schematic diagram of the original image collected by the grab camera of the closed-loop anti-sway system of the grab ship unloader of the present invention.

[0030] Figure 5 The schematic diagram is a diagram of marking the grab bucket of the original image collected by the closed-loop anti-sway system of the grab-type ship unloader of the present invention.

[0031] Figure 6 It is a schematic diagram of the closed-loop anti-sway system of the grab-type ship unloader of the present invention outputting coordinate values ​​of the detected grab.

[0032] In the above drawings, the reference numerals used are as follows:

[0033] 100 Closed loop anti-sway system for grab ship unloader

[0034] 101 Grab Ship Unloader

[0035] 102 Car

[0036] 103 Grab

[0037] 104 Main beam

[0038] 105 Discharge frame

[0039] 106 Electrical Room

[0040] 107 Rope

[0041] 10 Grab camera

[0042] 20 Video Processing Server

[0043] 21 Preliminary processing module

[0044] 22 Training Modules

[0045] 23 Detection module

[0046] 30 Anti-sway controller

[0047] 40 Crane main controller

[0048] 201 Cargo Ship

[0049] 202 Cargo DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described in detail with reference to the following embodiments.

[0051] Figure 1 Schematic diagram of the closed-loop anti-sway system of a grab ship unloader of the present invention applied to a grab ship unloader, Figure 2 This is a schematic diagram of the closed-loop anti-sway system of the grab ship unloader of the present invention. Figure 1 and Figure 2The closed-loop anti-sway system 100 of the grab ship unloader of the present invention is applied to the grab ship unloader 101. The grab ship unloader 101 is used to unload the cargo ship 201 moored at the dock. The cargo ship 201 has cargo 202 (such as coal, but not limited to this) in its cabin. The grab ship unloader 101 includes a trolley 102, a grab 103, a main beam 104, a discharge frame 105 and an electrical room 106. The grab 103 is connected to the trolley 102 by a rope 107. The trolley 102 can travel and move on the main beam 104. The discharge frame 105 is arranged below the main beam 104. After the grab 103 grabs the cargo 202, it is driven by the trolley 102 and the lifting mechanism, and the grab 103 discharges the cargo 202 into the discharge frame 105.

[0052] The closed-loop anti-sway system 100 of the grab ship unloader includes two grab cameras 10, a video processing server 20, an anti-sway controller 30 and a crane main controller 40. The two grab cameras 10 are arranged at both ends of the travel track of the trolley 102 on the main beam 104. The field of view of the grab camera 10 covers the movement range of the grab 103 of the grab ship unloader 101 during operation. The grab camera 10 collects video information of the grab ship unloader 101 during operation.

[0053] The video processing server 20, the anti-sway controller 30 and the crane main controller 40 are arranged in the electrical room 106. The video processing server 20 receives the video information from the grab bucket camera 10, extracts the grab bucket 103 from the video information, and generates the pixel coordinates of the grab bucket 103. The video processing server 20 is connected to the anti-sway controller 30 and the crane main controller 40. The video processing server 20 combines the pixel coordinates of the grab bucket 103 and the working condition data from the crane main controller 40 to calculate the actual space coordinates of the grab bucket 103. The anti-sway controller 30 obtains the coordinate information of the trolley 102, the operation instruction information of the trolley 102, the length information of the rope 107, and the operation instruction information of the rope 107 from the crane main controller 40, and calculates the real-time additional speed of the trolley 102 and the real-time additional speed of the rope 107 based on this information and the actual spatial coordinates of the grab 103, and sends the real-time additional speed to the crane main controller 40. The crane main controller 40 adjusts the movement of the trolley 102 and the rope 107 according to the real-time additional speed so that the current position of the grab 103 is controlled within a preset range.

[0054] It should be noted that the operating data of the crane main controller 40 includes the real-time position coordinates of the trolley 102 (for range reference), the length coordinates of the rope 107 (swing length), the opening and closing posture of the grab bucket 103 and the real-time load capacity of the grab bucket 103 .

[0055] More specifically, Figure 3 for Figure 2 The schematic diagram of the architecture of the video processing server of the closed-loop anti-sway system of the grab ship unloader is shown in FIG. Figure 4 This is a schematic diagram of the grab camera collecting the original image of the closed-loop anti-sway system of the grab ship unloader of the present invention. Figure 5 The schematic diagram of the grab bucket of the original image collected is marked by the closed-loop anti-sway system of the grab bucket ship unloader of the present invention. Figure 6 This is a schematic diagram of the grab-type ship unloader closed-loop anti-sway system of the present invention outputting the coordinate values ​​of the detected grab. Please also refer to Figures 3 to 6 The video processing server 20 includes a preliminary processing module 21, a training module 22 and a detection module 23. The preliminary processing module 21 receives video information from the grab bucket camera 10, extracts different frames in different video information at intervals to obtain more than one thousand original frame images for training the algorithm, and annotates various postures of the grab bucket 103 by calibration to generate an annotation file.

[0056] The training module 22 organizes the labeled files into a deep learning training set, and then uses the deep learning training set as the true value to execute the training algorithm to obtain a training result weight file.

[0057] The detection module 23 receives real-time video information and uses the result weight file to extract the grab 103 from the video information and generate pixel coordinates of the grab 103. The actual spatial coordinates of the grab 103 are calculated by combining the pixel coordinates of the grab 103 and the operating data from the crane main controller 40 (including the real-time position coordinates of the trolley 102, the length coordinates of the rope 107, the opening and closing posture of the grab 103 and the real-time load of the grab 103).

[0058] It should be noted that each frame in the real-time video of the grab bucket 103 is transmitted to the video detection module 23 through the network, and the predicted target structure frame (i.e., the grab bucket 103) is obtained, and finally the predicted structure frame is extracted from the video to generate pixel coordinates, and the actual spatial coordinates of the grab bucket 103 are calculated in combination with the working condition data input by the crane main controller 40. The weight file is a fixed algorithm calculated from the historical video frames, and the video processing server 20 uses the real-time frame data as an input parameter, and finally generates a real-time output result through calculation with the weight file.

[0059] After the grab bucket 103 is full of cargo 202, the crane main controller 40 controls the rope 107 to perform lifting movement first, and then controls the rope 107 and the trolley 102 to move simultaneously. The lifting rope 107 can be telescopically controlled and the trolley 102 can travel on the main beam 104. When the grab bucket 103 is located above the unloading frame 105, the grab bucket 103 unloads the cargo 202 into the unloading frame 105.

[0060] The present invention further provides a grab ship unloader 101, which comprises the above-mentioned grab ship unloader closed-loop anti-sway system 100.

[0061] The present invention also provides a closed-loop anti-sway method for a grab ship unloader, which comprises the following steps:

[0062] Collecting video information of the grab-type ship unloader 101 during operation, wherein the field of view of the video information covers the movement range of the grab 103 of the grab-type ship unloader 101 during operation;

[0063] Extract the grab bucket 103 from the video information of the grab bucket camera 10, generate the pixel coordinates of the grab bucket 103, and calculate and obtain the actual space coordinates of the grab bucket 103 by combining the pixel coordinates of the grab bucket 103 and the working condition data;

[0064] According to the motion information of the trolley 102 and the actual space coordinates of the grab bucket 103, the additional speed of the trolley 102 is determined, and the motion of the trolley 102 is adjusted according to the additional speed so that the current position of the grab bucket 103 is controlled within a preset range.

[0065] The closed-loop anti-sway method for a grab ship unloader of the present invention further comprises the following steps:

[0066] Receive video information from the grab bucket camera 10, extract different frames in different video information at intervals to obtain more than one thousand original frame images for training the algorithm, and mark various postures of the grab bucket 103 by calibration to generate a marking file;

[0067] Arranging the annotated files into a deep learning training set, and then executing a training algorithm using the deep learning training set as a true value to obtain a training result weight file; and

[0068] Receive real-time video information, and use the training result weight file to extract the grab bucket 103 from the real-time video information and generate pixel coordinates of the grab bucket 103.

[0069] The present invention also proposes an edge computing device, which includes: at least one memory and at least one processor; the at least one memory is used to store a machine-readable program; the at least one processor is used to call the machine-readable program to execute the above-mentioned closed-loop anti-sway method of the grab ship unloader.

[0070] The present invention further proposes a computer-readable medium, characterized in that computer instructions are stored on the computer-readable medium, and when the computer instructions are executed by a processor, the processor executes the above-mentioned closed-loop anti-sway method for a grab ship unloader.

[0071] The present invention also provides a computer program, comprising computer executable instructions, which, when executed, enable at least one processor to execute the above-mentioned closed-loop anti-sway method for a grab-type ship unloader.

[0072] The present invention further proposes a computer program product, which is tangibly stored on a computer-readable medium and includes computer-executable instructions, which, when executed, enable at least one processor to perform the above-mentioned closed-loop anti-sway method for a grab-type ship unloader.

[0073] The closed-loop anti-sway system of a grab ship unloader, the closed-loop anti-sway method of a grab ship unloader, the grab ship unloader, the edge computing device, the computer-readable medium, the computer program and the computer program product of the present invention have at least the following advantages:

[0074] 1. In the closed-loop anti-sway system, closed-loop anti-sway method, grab ship unloader, edge computing device, computer-readable medium, computer program and computer program product of the grab ship unloader of the present invention, video information of the grab ship unloader during operation is collected, the grab is extracted from the video information, and the pixel coordinates of the grab are generated. The actual space coordinates of the grab are calculated by combining the pixel coordinates of the grab and the working condition data. Then, the real-time additional speed of the trolley and the real-time additional speed of the rope are calculated according to the coordinate information and operation instruction information of the trolley, the length information and operation instruction information of the rope and the actual space coordinates of the grab. The movement of the trolley and the rope is adjusted according to the real-time additional speed so that the current position of the grab is controlled within a preset range, thereby effectively preventing the grab from shaking during operation. The present invention can achieve lower production costs and maintenance costs.

[0075] 2. In one embodiment of the closed-loop anti-sway system, closed-loop anti-sway method, grab ship unloader, edge computing device, computer-readable medium, computer program and computer program product of the present invention, video information of the entire operation process of the grab is captured in real time, and a large amount of actual videos collected are manually labeled and deep learning trained through visual deep learning technology to realize a customized grab extraction algorithm, which is then applied to the processing of real-time video, thereby realizing real-time position capture and redundant calculation of the grab, and realizing precise positioning of the grab.

[0076] 3. In one embodiment of the grab ship unloader closed-loop anti-sway system, grab ship unloader closed-loop anti-sway method, grab ship unloader, edge computing device, computer-readable medium, computer program and computer program product of the present invention, the mature artificial intelligence technology in visual processing is successfully applied to the field of traditional cranes, effectively solving the key difficulties in the anti-sway control of the ship unloader, with low hardware cost and mechanical modification cost, realizing closed-loop control of the grab, high detection accuracy, good system stability, simple structure, convenient construction and flexible application.

[0077] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A closed-loop anti-sway system (100) for a grab-type ship unloader, applied to a grab-type ship unloader (101), wherein the grab-type ship unloader (101) comprises a trolley (102), a grab (103) and a main beam (104), wherein the grab (103) is located below the trolley (102) and connected to the trolley (102) via a rope (107), and the trolley (102) can travel and move on the main beam (104), characterized in that: include: Two grab bucket cameras (10) are arranged at two ends of the travel track of the trolley (102) on the main beam (104); the field of view of the grab bucket cameras (10) covers the movement range of the grab bucket (103) during the operation of the grab bucket ship unloader (101); and the grab bucket cameras (10) collect video information during the operation of the grab bucket ship unloader (101); A video processing server (20) receives video information from the grab bucket camera (10), extracts the grab bucket (103) from the video information, and generates pixel coordinates of the grab bucket (103); An anti-sway controller (30); and Crane main controller (40); The video processing server (20) is connected to the anti-sway controller (30) and the crane main controller (40); the video processing server (20) calculates the actual spatial coordinates of the grab (103) in combination with the pixel coordinates of the grab (103) and the working condition data from the crane main controller (40); the anti-sway controller (30) obtains the coordinate information and operation instruction information of the trolley (102) and the length information and operation instruction information of the rope (107) from the crane main controller (40), and calculates the real-time additional speed of the trolley (102) and the real-time additional speed of the rope (107) based on these information and the actual spatial coordinates of the grab (103), and sends the real-time additional speed to the crane main controller (40); the crane main controller (40) adjusts the movement of the trolley (102) and the rope (107) according to the real-time additional speed, so that the current position of the grab (103) is controlled within a preset range.

2. The closed-loop anti-sway system (100) of grab ship unloader according to claim 1, characterized in that: The video processing server (20) comprises: A preliminary processing module (21) receives video information from the grab bucket camera (10), extracts different frames in different video information at intervals to obtain more than one thousand original frame images for training the algorithm, and performs target annotation on various postures of the grab bucket (103) by calibration to generate an annotation file; A training module (22) organizes the annotated files into a deep learning training set, and then uses the deep learning training set as a true value to execute a training algorithm to obtain a training result weight file; and A detection module (23) receives real-time video information and uses the training result weight file to extract the grab (103) from the video information and generate pixel coordinates of the grab (103). The actual spatial coordinates of the grab (103) are calculated by combining the pixel coordinates of the grab (103) and the working condition data from the crane main controller (40).

3. A grab ship unloader (101), characterized in that: It comprises the closed-loop anti-sway system (100) of the grab ship unloader as claimed in claim 1.

4. The grab ship unloader according to claim 3, characterized in that: The video processing server (20) comprises: A preliminary processing module (21) receives video information from the grab bucket camera (10), extracts different frames in different video information at intervals to obtain more than one thousand original frame images for training the algorithm, and performs target annotation on various postures of the grab bucket (103) by calibration to generate an annotation file; A training module (22) organizes the annotated files into a deep learning training set, and then uses the deep learning training set as a true value to execute a training algorithm to obtain a training result weight file; and A detection module (23) receives real-time video information and uses the training result weight file to extract the grab (103) from the real-time video information and generate pixel coordinates of the grab (103). The actual spatial coordinates of the grab (103) are calculated by combining the pixel coordinates of the grab (103) and the working condition data from the crane main controller (40).

5. A closed-loop anti-sway method for a grab-type ship unloader, applied to a grab-type ship unloader (101), wherein the grab-type ship unloader (101) comprises a trolley (102), a grab (103) and a main beam (104), wherein the grab (103) is located below the trolley (102) and connected to the trolley (102) via a rope (107), and the trolley (102) can travel and move on the main beam (104), and the closed-loop anti-sway method for the grab-type ship unloader comprises the following steps: Collecting video information of the grab-type ship unloader (101) during operation, wherein the field of view of the video information covers the movement range of the grab (103) of the grab-type ship unloader (101) during operation; Extracting the grab (103) from the video information, generating pixel coordinates of the grab (103), and calculating and obtaining the actual spatial coordinates of the grab (103) by combining the pixel coordinates of the grab (103) and working condition data; According to the coordinate information and operation instruction information of the trolley (102), the length information and operation instruction information of the rope (107) and the actual space coordinates of the grab bucket (103), the real-time additional speed of the trolley (102) and the real-time additional speed of the rope (107) are calculated, and the movement of the trolley (102) and the rope (107) is adjusted according to the real-time additional speed so that the current position of the grab bucket (103) is controlled within a preset range.

6. The closed-loop anti-sway method for grab ship unloader according to claim 5, characterized in that: The following steps are also included: Receiving video information from the grab bucket camera (10), extracting different frames in different video information at intervals to obtain more than one thousand original frame images for training the algorithm, and marking various postures of the grab bucket (103) by calibration to generate a marking file; Arrange the annotated files into a deep learning training set, and then use the deep learning training set as a true value to execute a training algorithm to obtain a training result weight file; as well as Receive real-time video information, and use the training result weight file to extract the grab bucket (103) from the real-time video information and generate pixel coordinates of the grab bucket (103).

7. An edge computing device, characterized in that: The apparatus includes: at least one memory and at least one processor; The at least one memory is used to store a machine-readable program; The at least one processor is used to call the machine-readable program to execute the closed-loop anti-sway method for the grab ship unloader according to any one of claims 5-6.

8. A computer-readable medium, characterized in that The computer readable medium stores computer instructions, and when the computer instructions are executed by the processor, the processor executes the closed-loop anti-sway method for a grab ship unloader according to any one of claims 5-6.

9. A computer program, comprising computer executable instructions, which, when executed, enable at least one processor to execute the closed-loop anti-sway method for a grab ship unloader according to any one of claims 5 to 6.

10. A computer program product, which is tangibly stored on a computer-readable medium and includes computer-executable instructions, which when executed, enable at least one processor to perform the closed-loop anti-sway method for a grab ship unloader according to any one of claims 5 to 6.

Images