Device and method for automatically identifying multiplying power of steel wire rope of crane

Abstract

The invention discloses a device and method for automatically identifying the multiplying power of a crane steel wire rope, and the device comprises an image collection device which carries out the image collection of a crane hook and a pulley block, and an edge calculation unit which carries out the intelligent analysis and preprocessing through inputting an image into a trained multiplying power identification deep learning model, and identifies the actual multiplying power of a crane. And the display is used for outputting the multiplying power result of the crane steel wire rope. The position of the steel wire rope can be determined through the image recognition method by shooting the pictures of the lifting hook and the lifting hook pulley block through the camera, then the straightened steel wire rope image is obtained to automatically recognize the number of strands, the lifting hook type can be automatically recognized through the system, and meanwhile the number of strands of the steel wire rope is automatically recognized to calculate the multiplying power. And the identification device can determine the multiplying power without depending on a special sensor, so that the manual operation error input rate or the installation cost of the special identification device is reduced.

Description

Technical Field

[0001] The present invention relates to engineering machinery, and in particular to a device and method for automatically identifying the ratio of a crane wire rope. Background Art

[0002] Crane hoisting operations require selecting different operating conditions and setting the corresponding wire rope ratio. The ratio refers to the force-saving or speed-increasing ratio of the crane's wire rope pulley assembly. This ratio is determined by the number of strands of wire rope between the fixed and movable pulleys, thus calculating the crane's actual ratio. Traditionally, ratio detection relies on manual observation, requiring close observation to clearly determine the strand count at long distances or in adverse environmental conditions. Wire rope ratio is a crucial parameter, and automated identification allows for more efficient and safer operation.

[0003] The existing technology installs multiple detection switches and photoelectric switches on the pulley block and the boom, and the multiple detection switches and photoelectric switches are connected to the magnification recognition device. The magnification recognition device calculates the magnification of the wire rope based on the number of boom pulleys around which the wire rope is wrapped and the fixed point at the end of the wire rope.

[0004] Existing technology relies on multiple detection switches and photoelectric switches connected to a magnification recognition device for magnification identification. This requires installing a detection switch for each corresponding boom pulley, and for each pulley set, making installation complex and costly. The photoelectric switches used are expensive, so many cranes don't use them. Furthermore, the identification device, which includes a signal collector on the boom, a receiver in the cab, and a torque limiter, requires complex connections with other switches, potentially causing damage in harsh lifting environments.

[0005] Therefore, it is necessary to develop a new device and method for automatically identifying the crane wire rope ratio to overcome the problems of the existing technology. Summary of the Invention

[0006] Purpose of the invention: In view of the shortcomings and defects of the existing technology, the present invention provides a device and method for automatically identifying the ratio of crane wire ropes. The position of the wire rope can be determined by using an image recognition method by taking pictures of the hook and the hook pulley assembly with a camera, and then the number of strands can be automatically identified by obtaining the image of the straightened wire rope. The system can automatically identify the type of hook and automatically identify the number of strands of the wire rope to calculate the ratio. The ratio can be determined without relying on special sensors or identification devices, thereby reducing the error input rate of manual operation or the installation cost of special identification devices.

[0007] Technical solution: The present invention provides a device for automatically identifying the crane wire rope ratio, which is characterized by: including an image acquisition device for capturing images of the crane hook and pulley assembly, an edge computing unit for intelligent analysis and preprocessing of the image input into a trained ratio recognition deep learning model, and identifying the actual ratio of the crane, and a display for outputting the results of the crane wire rope ratio.

[0008] Wherein, the image acquisition device is an industrial camera, which is arranged on a turntable and faces the hook pulley block at the end of the boom.

[0009] The present invention provides a method for automatically identifying the ratio of a crane wire rope, which is characterized in that: an image acquisition device is installed on the turntable of the crane, the position is identified through the hook and pulley group, the camera adjusts the viewing angle upward according to the detected position to ensure that the wire rope above the hook is photographed, and the collected wire rope sample pictures are used to establish an early ratio recognition data set; the ratio recognition database is established by marking the wire rope above the hook to input the wire rope ratio recognition deep learning model for training, and the training result can automatically identify the wire rope in the input hook wire rope picture; the wire rope is identified as a strand of the wire rope only when it passes through the corresponding hook and pulley group; the actual ratio is obtained by calculation of the deep learning model and transmitted to the display.

[0010] Among them, when the aforementioned early rate recognition data set is established, sample images of different types of hooks and pulley groups under different crane operating scenarios are collected to establish a hook model recognition training set. The training model automatically identifies the hook model and provides a judgment standard for subsequent rate self-inspection.

[0011] Among them, the wire rope ratio recognition deep learning model is a neural network based on computer vision, including a backbone network Backbone, a feature fusion network Neck and a detection head Head.

[0012] The wire rope ratio recognition deep learning model is trained not only on wire ropes but also on images of all crane hooks. Different hooks are categorized and labeled. This means the trained model can identify hook models, determine the maximum number of wire rope strands that can pass through them, and then calculate the actual crane ratio using images of the pulley and wire rope. Finally, the calculated actual ratio is transmitted to a display screen via a communication module for visualization.

[0013] The system automatically checks the wire rope ratio based on the detected crane hook type. If the identified wire rope strand count exceeds the maximum threading capacity for that hook type, the identification is invalid. A human-machine interface alarm alerts the operator, requiring manual confirmation or re-automatic ratio identification until the ratio is correctly identified and potential operational risks are eliminated. This real-time adjustment helps improve crane operation safety.

[0014] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: the present invention uses a camera to take pictures of the hook and the hook pulley assembly and relies on image recognition methods to determine the position of the wire rope, and then obtains the image of the straightened wire rope to automatically identify the number of strands. The system can automatically identify the type of hook and automatically identify the number of wire rope strands to calculate the magnification. The magnification can be determined without relying on special sensors or recognition devices, thereby reducing the error input rate of manual operation or the installation cost of special recognition devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a structural schematic diagram of the present invention;

[0016] Figure 2 This is a schematic diagram of the installation structure of the industrial camera of the present invention;

[0017] Figure 3 This is a logic flow chart of the magnification recognition of the present invention;

[0018] In the figure, 1 is the hook pulley assembly; 2 is the boom; 3 is the industrial camera; and 4 is the turntable. DETAILED DESCRIPTION

[0019] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and specific implementation methods.

[0020] The present invention's device for automatically identifying crane rope magnifications includes an image acquisition device that captures images of the crane hook and pulley assembly; an edge computing unit that intelligently analyzes and preprocesses the images using a trained deep learning model for magnification recognition, identifying the crane's actual magnification; and a display that outputs the crane rope magnification results. The image acquisition device is an industrial camera 3 mounted on a turntable 4, facing the hook and pulley assembly 1 at the end of the boom 2.

[0021] The present invention provides a method for automatically identifying the ratio of a crane wire rope. The image acquisition device is installed on the crane's turntable. The position is identified through the hook and pulley assembly. The camera adjusts the viewing angle upward according to the detected position to ensure that the wire rope above the hook is captured. The collected wire rope sample images are used to establish an early ratio recognition data set. The ratio recognition database is established by marking the wire rope above the hook to input the wire rope ratio recognition deep learning model for training. The training result can automatically identify the wire rope in the input hook wire rope image. Only when the wire rope passes through the corresponding hook and pulley assembly is it identified as a strand of the wire rope. The actual ratio is calculated by the deep learning model and transmitted to the display. When establishing the early ratio recognition data set, sample images of different types of hooks and pulley assemblies under different crane operating scenarios are collected to establish a hook model recognition training set. The training model automatically identifies the hook model and provides a judgment standard for subsequent ratio self-test. The deep learning model for wire rope ratio recognition is a neural network based on computer vision, including a backbone network, a feature fusion network, and a detection head. The typical model is based on the YOLO series model.

[0022] In addition to training on wire ropes, the deep learning model for wire rope ratio recognition also trains on images of all crane hooks. Different hooks are categorized and labeled. After training, the model can identify hook models, determine the maximum number of wire rope strands that can be threaded through the hook model, and then calculate the actual crane ratio using images of the pulley block and wire rope. Finally, the actual ratio is calculated and transmitted to the display screen via the communication module for visualization. The system performs a self-check on the identified wire rope ratio based on the detected crane hook type. If the identified wire rope strand count exceeds the maximum threading capacity for that hook type, the recognition is invalid. An alarm is issued to the operator through the human-machine interface, requiring manual confirmation or re-trying the automatic ratio recognition until the ratio is correctly identified and no potential operational risks are identified. These real-time adjustments help improve crane operation safety.

[0023] The deep learning target detection algorithm of the present invention can add data enhancements such as random occlusion and motion blur to better train the model and improve the accuracy of image recognition; establish a multi-working condition training data set (including various types of cranes and hook groups); and deploy a lightweight model to compress parameters.

[0024] The present invention uses a camera to take pictures of the hook and the hook pulley assembly and relies on an image recognition method to determine the position of the wire rope, and then obtains the image of the straightened wire rope to automatically identify the number of strands. Relying on this system, the hook type can be automatically identified, and the number of strands of the wire rope can be automatically identified to calculate the magnification. The magnification can be determined without relying on special sensors or recognition devices, thereby reducing the error input rate of manual operation or the installation cost of special recognition devices.

Claims

1. A device for automatically identifying the crane wire rope ratio, characterized by: It includes an image acquisition device that captures images of the crane hook and pulley assembly, an edge computing unit that performs intelligent analysis and preprocessing on the image input into a trained magnification recognition deep learning model, identifies the actual magnification of the crane, and a display that outputs the results of the crane wire rope magnification.

2. The device for automatically identifying the crane wire rope ratio according to claim 1, characterized in that: The image acquisition device is an industrial camera (3), which is arranged on a turntable (4) and faces a hook pulley assembly (1) at the end of a boom (2).

3. A method for automatically identifying the crane wire rope ratio, characterized by: The image acquisition device is installed on the turntable of the crane. The position is identified through the hook and pulley. The camera adjusts the viewing angle upward according to the detected position to ensure that the wire rope above the hook is captured. The collected wire rope sample images are used to establish the early magnification recognition data set. The method for establishing the rate recognition database is to input the wire rope rate recognition deep learning model for training by marking the wire rope above the hook. The training result can automatically identify the wire rope in the input hook and wire rope image; when the wire rope passes through the corresponding hook and pulley group, it is identified as a strand of wire rope; The actual magnification is obtained through calculation of the deep learning model and transmitted to the display.

4. The method for automatically identifying the crane wire rope ratio according to claim 3, characterized in that: When establishing the aforementioned early rate recognition data set, sample images of different types of hooks and pulleys in different crane operating scenarios are collected to establish a hook model recognition training set. The training model automatically identifies the hook model and provides a judgment standard for subsequent rate self-inspection.

5. The method for automatically identifying the crane wire rope ratio according to claim 3, characterized in that: The wire rope ratio recognition deep learning model is a neural network based on computer vision, including a backbone network Backbone, a feature fusion network Neck and a detection head Head.

6. The method for automatically identifying the crane wire rope ratio according to claim 5, characterized in that: In addition to training on wire ropes, the wire rope ratio recognition deep learning model also trains on images of all crane hooks and labels different hooks. That is, after training, the model can identify the hook model, determine the maximum number of wire rope strands that can pass through the hook model, and then calculate the actual crane ratio using the pulley block and wire rope images. Finally, the actual ratio is calculated and transmitted to the display screen by the communication module for visualization.

7. The method for automatically identifying the crane wire rope ratio according to claim 5, characterized in that: The system performs a self-check on the identified wire rope ratio based on the detected crane hook type. If the identified wire rope strand count is greater than the maximum rope threading capacity for this type of hook, the recognition is invalid. The operator is prompted with an alarm through the human-machine interface, requiring manual confirmation or re-automatic ratio recognition until the ratio is correctly identified and there are no potential operational risks, thereby improving the safety of crane operations.

Images