Overhead Crane Operation Control Method and Overhead Crane Equipment

Abstract

The present invention discloses a method for controlling the operation of an overhead crane and an overhead crane device. In the method for controlling the operation of an overhead crane, during the movement of an overhead crane, an image collected by a vision detection device on the overhead crane is obtained in real time, and a traffic environment within a certain range in front of the overhead crane is determined based on the image collected by the vision detection device on the overhead crane obtained in real time and the information obtained from a scheduling system; at least according to the traffic environment within a certain range in front of the overhead crane, it is determined whether the overhead crane where the vision detection device is located needs to enter a protection operation mode; when entering the protection operation mode, the moving speed of the vision detection device is controlled based on the real-time distance between the overhead crane where the vision detection device is located determined based on the image collected by the vision detection device obtained in real time and another overhead crane moving in the same direction in front of it. Compared with traditional radar or laser positioning, the present invention has higher accuracy and reliability through a vision solution.

Description

Technical Field

[0001] The invention relates to the field of automatic material handling systems, in particular to an overhead crane operation control method and an overhead crane device. Background Art

[0002] The OHT (Overhead Hoist Transfer) system, also known as the overhead crane handling system, is widely used in automated factories such as semiconductor manufacturing.

[0003] In an OHT system, there will be multiple overhead cranes running at the same time. Therefore, anti-collision between overhead cranes in the OHT system is very important.

[0004] The commonly used anti-collision method is to perform anti-collision control through radar or laser positioning, but in the OHT system, the radar or laser will be subject to more interference, there will be certain misjudgments, affecting the operating efficiency of the overhead crane handling system. Summary of the invention

[0005] The purpose of the present invention is to solve the above problems existing in the prior art and to provide an overhead crane operation control method and an overhead crane device.

[0006] The purpose of the present invention is achieved through the following technical solutions:

[0007] An overhead crane operation control method, during the movement of an overhead crane, acquires images captured by a visual detection device on the overhead crane in real time, and determines a traffic environment within a certain range in front of the overhead crane based on the images captured by the visual detection device on the overhead crane acquired in real time and information acquired from a dispatching system;

[0008] Determining whether the overhead crane where the visual inspection device is located needs to enter a protective operation mode at least according to the traffic environment within a certain range in front of the overhead crane;

[0009] When entering the protection operation mode, the moving speed of the visual inspection device is controlled based on the real-time distance between the overhead crane where the visual inspection device is located and another overhead crane in front of it moving in the same direction, which is determined based on the image captured by the visual inspection device acquired in real time.

[0010] Preferably, the traffic environment within a certain range in front of the overhead crane is one of the following situations, including no overhead crane, another overhead crane traveling in the same direction as the overhead crane and maintaining a normal moving speed, another overhead crane traveling in the same direction as the overhead crane and decelerating, another overhead crane traveling straight in the same direction as the overhead crane and accelerating, another overhead crane stopped on its moving route, and another overhead crane turning on its moving route.

[0011] Preferably, determining the traffic environment within a certain range in front of the overhead crane based on the images collected by the visual detection device on the overhead crane obtained in real time and the information obtained from the dispatching system includes the following process:

[0012] Input the images collected by the visual detection device obtained in real time into the trained target detection model after being subjected to predetermined processing to determine the situation of the overhead crane within a certain range in front of the overhead crane, and the situation of the overhead crane is one of no overhead crane, there is another overhead crane and it is traveling in the same direction, and there is another overhead crane turning on its moving route;

[0013] When it is determined that the situation of the overhead crane within a certain range in front of the overhead crane is no overhead crane or there is another overhead crane turning on its moving route, it is determined that the traffic environment within a certain range in front of the overhead crane is no overhead crane or there is another overhead crane turning on its moving route;

[0014] When it is determined that the situation of the overhead crane within a certain range in front of the overhead crane is there is another overhead crane and it is traveling in the same direction, then determine the traffic environment within a certain range in front of the overhead crane according to the speed situation of the other overhead crane obtained from the dispatching system, which is that there is another overhead crane and it is traveling in the same direction and decelerating, or there is another overhead crane and it is going straight in the same direction and accelerating, or there is another overhead crane stopping on its moving route, or there is another overhead crane and it is going straight in the same direction and maintaining a normal moving speed.

[0015] Preferably, the process of subjecting the images collected by the visual detection device obtained in real time to predetermined processing includes:

[0016] Perform image enhancement processing on the images collected by the two cameras respectively to obtain an enhanced right view image and an enhanced left view image;

[0017] Construct image pyramids corresponding to the enhanced right view image and the enhanced left view image respectively and obtain a left image feature map and a right image feature map based on the image pyramids;

[0018] Based on the left image feature map and the right image feature map, perform preliminary disparity estimation at the lowest resolution layer to obtain a disparity map;

[0019] Obtain a compensated disparity map after compensating the disparity map.

[0020] Preferably, the image enhancement processing includes the following process: First, perform histogram equalization on the image, then perform Laplacian operator enhancement, and finally perform Gamma correction.

[0021] Preferably, the successive compensation of the disparity map includes structured light compensation, temperature compensation, and vibration compensation.

[0022] Preferably, if it is determined that the passing environment within a certain range in front of the overhead crane has another overhead crane traveling in the same direction and decelerating, another overhead crane stopping on its moving route, or another overhead crane turning on its moving route, it is determined that the overhead crane where the visual detection device is located needs to enter the protection operation mode.

[0023] Preferably, the overhead crane where the visual detection device is located receives in real time the running track of the other overhead crane fed back by the dispatching system. If it is determined that the passing environment within a certain range in front of the overhead crane has another overhead crane traveling in the same direction and maintaining a normal moving speed, and the running track of the other overhead crane fed back by the dispatching system is about to turn, it is determined that the overhead crane where the visual detection device is located needs to enter the protection operation mode.

[0024] Preferably, the overhead crane where the visual detection device is located receives in real time the running tracks of other overhead cranes fed back by the dispatching system. If it is determined that another overhead crane and the overhead crane where the visual detection device is located are on two intersecting lines, and the other overhead crane needs to turn at the intersection of the two lines, if it is determined that the overhead crane where the visual detection device is located is going to go straight through the intersection, when the visual detection device on the overhead crane detects the other overhead crane that is going to turn at the intersection, the other overhead crane is stopped from moving, and it is determined that the overhead crane where the visual detection device is located needs to enter the protection operation mode;

[0025] If it is determined that the overhead crane where the visual detection device is located is going to turn at the intersection, when the visual detection device on the overhead crane detects the other overhead crane that is going to turn at the intersection, the overhead crane where the visual detection device is located enters the protection operation mode, and the other overhead crane is made to turn through the intersection first.

[0026] Preferably, the real-time distance is determined according to the following formula:

[0027] Z = (b * f) / d;

[0028] Wherein, Z is the real-time distance, b is the baseline length, f is the camera focal length, and d is the parallax.

[0029] Preferably, in the protection operation mode, the moving speed of the overhead crane where the visual detection device is located is controlled according to the following process;

[0030] Determine whether the real-time distance is less than the deceleration distance;

[0031] If not, make the overhead crane where the visual detection device is located travel at the normal moving speed and repeat determining whether the real-time distance reaches the deceleration distance;

[0032] If so, make the overhead crane where the visual detection device is located decelerate and determine the change situation of the real-time distance;

[0033] If it is determined that the real-time distance is greater than the speed reduction distance, the overhead travelling crane where the visual inspection device is located is restored to a normal moving speed;

[0034] If it is determined that the real-time distance is between the deceleration distance and the parking distance, the overhead travelling vehicle where the visual inspection device is located continues to decelerate or maintains the current moving speed;

[0035] If it is determined that the real-time distance is less than the parking distance, the moving speed of the overhead crane where the visual inspection device is located is controlled to 0, and it is continued to be determined whether the real-time distance exceeds the parking distance. If so, the overhead crane where the visual inspection device is located is accelerated and the change of the real-time speed is determined.

[0036] An overhead crane device comprises a memory and a processor, wherein the memory stores a program executable by the processor, and when the program is executed, any of the above-mentioned overhead crane operation control methods is implemented.

[0037] The advantages of the technical solution of the present invention are mainly reflected in:

[0038] The present invention uses a visual solution, and the overhead crane can use the image information captured by the camera, combined with image processing technology and algorithms, to achieve accurate recognition and positioning of an overhead crane in front of it. This positioning method has higher accuracy and reliability than traditional radar or laser positioning. The use of a visual solution can also achieve dynamic tracking of moving targets, ensuring that the overhead crane can accurately follow the target path during transportation, reducing offsets and errors. At the same time, corresponding obstacle avoidance processing can be performed based on higher recognition accuracy, thereby ensuring the safety of the transportation process and adopting a step-by-step speed regulation method to be conducive to balancing operating efficiency.

[0039] The visual solution of the present invention can realize multi-dimensional information acquisition, including color, shape, texture, etc. This information helps the overhead crane to perceive the surrounding environment more comprehensively and thus identify obstacles more accurately.

[0040] The present invention combines artificial intelligence technologies such as machine learning and deep learning, and the visual solution can also achieve autonomous learning and evolution, continuously improving its own perception, recognition and decision-making capabilities, making the overhead crane more intelligent.

[0041] The image processing method of the present invention can effectively improve the image quality, thereby creating favorable conditions for model training and subsequent image recognition, and is beneficial to improving the accuracy of image recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a schematic diagram of two image acquisition devices arranged on the overhead travelling crane of the present invention;

[0043] Figure 2It is a process schematic diagram of the method of the present invention;

[0044] Figure 3 It is a calculation schematic diagram for determining the real-time distance between another overhead crane and the overhead crane where the visual detection device is located in the present invention;

[0045] Figure 4 It is a process schematic diagram of the protection operation mode in the present invention. Detailed implementation manners

[0046] The objectives, advantages and features of the present invention will be illustrated and explained by the following non-restrictive description of preferred embodiments. These embodiments are only typical examples of applying the technical solutions of the present invention, and any technical solutions formed by equivalent replacement or equivalent transformation fall within the scope of protection required by the present invention.

[0047] In the description of the solution, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "front", "rear", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of description and simplification of the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, the terms "first", "second", "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0048] Embodiment 1

[0049] The overhead crane operation control method disclosed by the present invention will be described below with reference to the accompanying drawings. As shown in the attached drawings, the method of the present invention is based on setting a visual detection device on the forward side when each overhead crane 100 moves. The visual detection device preferably includes two image acquisition devices 200 arranged at the same height. The image acquisition device 200 can be a known camera. The two image acquisition devices 200 simulate the visual effect of human binoculars by taking left and right images of the same target, so as to obtain the three-dimensional information of the target. At the same time, a specific filter is installed in front of the lens of the camera to filter out unnecessary light, such as ambient light and infrared light, which is beneficial to subsequent image recognition. Figure 1 During the process of each overhead crane performing the handling task, the passing environment within a certain range in front of each overhead crane during its movement is analyzed in real time based on the images collected by the two cameras, so as to accurately control the operation of each overhead crane.

[0050] Specifically, as shown in the attached drawings

[0051] Specifically, as shown in the attached drawings Figure 2As shown, during the movement of a gantry crane in one day, images collected by the visual detection device on the gantry crane are obtained in real time, and the traffic environment within a certain range in front of the gantry crane is determined based on the images collected by the visual detection device on the gantry crane obtained in real time and the information obtained from the dispatching system;

[0052] Determine whether the gantry crane where the visual detection device is located needs to enter the protection operation mode at least according to the traffic environment within a certain range in front of the gantry crane;

[0053] When entering the protection operation mode, control the moving speed of the visual detection device based on the real-time distance between the gantry crane where the visual detection device is located determined from the images collected by the visual detection device obtained in real time and another gantry crane moving in the same direction in front of it.

[0054] The traffic environment within a certain range in front of the gantry crane includes no gantry crane, there is another gantry crane and it is moving in the same direction and maintaining a normal moving speed, there is another gantry crane and it is moving in the same direction and decelerating, there is another gantry crane and it is moving straight in the same direction and accelerating, there is another gantry crane stopped on its moving route, there is another gantry crane turning on its moving route. Here, there is another gantry crane turning on its moving route includes that another gantry crane has turned from the same moving direction to another moving direction and another gantry crane has turned from another moving direction to the same moving direction.

[0055] Determining the traffic environment within a certain range in front of the gantry crane based on the images collected by the visual detection device on the gantry crane obtained in real time and the information obtained from the dispatching system includes the following process:

[0056] The images collected by the visual detection device obtained in real time are input into the trained target detection model after being subjected to a predetermined process to determine the gantry crane situation within a certain range in front of the gantry crane. The gantry crane situation is one of no gantry crane, there is another gantry crane and it is moving in the same direction, there is another gantry crane turning on its moving route. Here, moving in the same direction means that the moving direction of another gantry crane is the same as the moving direction of the gantry crane where the visual detection device is located, and its moving route refers to the moving route of the gantry crane where the visual detection device is located; further, the gantry crane situation can also include whether there is another gantry crane within a certain range in front of the gantry crane, and the current track of the other gantry crane intersects with the current track of the gantry crane where the visual detection device is located.

[0057] When it is determined that the gantry crane situation within a certain range in front of the gantry crane is no gantry crane or there is another gantry crane turning on its moving route, then determine that the traffic environment within a certain range in front of the gantry crane is no gantry crane or there is another gantry crane turning on its moving route;

[0058] When it is determined that there is another overhead crane within a certain range in front of the overhead crane and it is moving in the same direction as it, the traffic environment within a certain range in front of the overhead crane is determined according to the speed condition of the other overhead crane obtained from the dispatching system, that is, there is another overhead crane moving in the same direction as it and decelerating, or there is another overhead crane moving straight in the same direction as it and accelerating, or there is another overhead crane stopped on its moving route, or there is another overhead crane moving straight in the same direction as it and maintaining a normal moving speed. The running speed of each overhead crane is fed back to the dispatching system in real time, so that the dispatching system knows the speed condition of each overhead crane, that is, knows whether each overhead crane is decelerating, accelerating, stopped, or moving at a constant speed. The dispatching system also knows the positions and moving trajectories of each overhead crane at the same time. Therefore, the dispatching system will feed back the speed condition, moving trajectory, etc. of the other overhead crane to the overhead crane where the vision detection device is located.

[0059] The target detection model can adopt the known SSD model, preferably the VGG16 model. Its specific construction method and training method are known technologies and will not be elaborated here. Regarding the image data required for model training, in the present invention, the images of one overhead crane running one lap are first collected in a real scene, then the images of two overhead cranes running one lap one after the other are collected, and then the collected images are subjected to a predetermined process before model training.

[0060] The predetermined process includes respectively performing image enhancement processing on the images collected by two cameras to obtain an enhanced right view image and an enhanced left view image; the image enhancement processing includes the following process: first, performing histogram equalization on the image, then performing Laplacian operator enhancement, and finally performing Gamma correction. The histogram equalization is used to adjust the gray distribution of the image, the Laplacian operator enhancement strengthens the edge part of the image through an edge detection operator, making the structural information in the image more obvious; the Gamma correction makes the dark part details clearer and the bright part not overexposed by adjusting the brightness of the image.

[0061] Then, an image pyramid corresponding to the enhanced right view image and the enhanced left view image is respectively constructed, and a left image feature map and a right image feature map are obtained based on the image pyramid; specifically, through the enhanced right view image and the enhanced left view image, layer-by-layer downsampling is performed to generate a series of image layers with different resolutions to obtain the image pyramid corresponding to the enhanced right view image and the enhanced left view image. Each image layer is a certain proportion of reduction of the previous image layer, so that the image can be analyzed and processed at multiple scales.

[0062] The obtained pyramid images are respectively input into a feasible convolutional neural network (such as the SSD model). Using the multi-layer convolutional neural network feature extraction method, through a combination of a series of convolutional layers, pooling layers, and activation functions, it can automatically learn and extract complex combinations in the images, thereby extracting left image feature maps and right image feature maps at different scales. Among them, the convolutional layer slides the convolution kernel (filter) on the image to extract the features of the local area; the pooling layer reduces the size of the feature map through downsampling operations while retaining important feature information; the activation function (such as ReLU) enables the network to learn more complex patterns through non-linear transformations.

[0063] Based on the left image feature map and the right image feature map, preliminary disparity estimation is performed at the lowest resolution layer to obtain the left image disparity map and the right image disparity map; here, the disparity estimation mainly adopts the block matching method. By searching for windows of a fixed size in the left image feature map and the right image feature map, the window pair with the highest similarity is found, and the disparity between them is calculated.

[0064] Specifically, it is calculated according to the following formula:

[0065]

[0066] where, d init (x,y) is the disparity map, d is the disparity value, representing the displacement difference of the pixel positions of the same target on the left image feature map and the right image feature map; w is the window size, and w represents the local neighborhood used to calculate the difference during the disparity estimation, defining the pixel range considered around each pixel; E is the lowest resolution layer, indicating the layer with the lowest resolution in the multi-scale pyramid; r is a preset weight coefficient, which ensures that the differences between the feature maps are reasonably processed in the calculation, making the matching process more accurate; F L (E) (x,y) is the image feature at the position with coordinates (x,y) in the left image feature map of the E-th layer; F R (E) (x,y) is the image feature at the position with coordinates (x,y) in the right image feature map of the E-th layer, F R (E) (x,y+d) is the image feature at the position with coordinates (x,y+d) in the right image feature map of the E-th layer.

[0067] The advantage of performing preliminary disparity estimation at the lowest resolution layer lies not only in high computational efficiency but also in its ability to provide a stable and highly robust initial disparity map.

[0068] After compensating the left image disparity map and the right image disparity map, the compensated left image disparity map and the right image disparity map are obtained and used as the corrected left and right image pair; the compensation for the left image disparity map and the right image disparity map includes structured light compensation, temperature compensation, and vibration compensation. Among them, the structured light compensation is to introduce a structured light pattern with a known pattern in the disparity map, such as a laser grid, stripes, or dot matrix. These structured light patterns can form distinct markers in the disparity map, and these markers will appear in both the left and right image pairs. The temperature compensation is to monitor the ambient temperature around the vision detection device in real time through a temperature sensor, combine the temperature response curve of the camera, and use digital signal processing methods to adjust the gray value and color balance of the disparity map. The vibration compensation is to deal with the small vibrations and displacements that the camera may suffer during the measurement process. The compensation can monitor the motion state of the camera in real time through a gyroscope and an acceleration sensor, and combine a known image stabilization algorithm to dynamically correct the image.

[0069] The corrected left and right image pair is input into the established target detection model for training until the model converges. After training the target detection model, the images collected in real time are processed according to the above-mentioned predetermined process to obtain the corrected left and right images and input them into the trained target detection model. The target detection model processes the left and right image pair into left and right regions of interest, adds classification labels to each frame of the image, performs stereo matching on the left and right image regions, and obtains a classification detection result, which is one of the situations of the overhead crane within a certain range in front of the overhead crane.

[0070] After determining the traffic environment within a certain range in front of the overhead crane, it is possible to determine whether the overhead crane where the vision detection device is located needs to enter the protection operation mode according to different traffic environments. Specifically:

[0071] If it is determined that the traffic environment within a certain range in front of the overhead crane is that there is another overhead crane traveling in the same direction and decelerating, there is another overhead crane stopping on its moving route, or there is another overhead crane turning on its moving route, it is determined that the overhead crane where the vision detection device is located needs to enter the protection operation mode.

[0072] Furthermore, while determining the traffic environment within a certain range in front of the overhead crane in real time, the overhead crane where the vision detection device is located also receives the running trajectory of the other overhead crane fed back by the dispatching system in real time. If it is determined that the other overhead crane and the overhead crane where the vision detection device is located are currently traveling in the same direction and the received running trajectory of the other overhead crane is about to turn, it is determined that the overhead crane where the vision detection device is located needs to enter the protection operation mode.

[0073] Meanwhile, the overhead crane where the visual detection device is located also receives in real time the running trajectories of other overhead cranes fed back by the dispatching system. The overhead crane where the visual detection device is located receives in real time the running trajectories of other overhead cranes fed back by the dispatching system. If it is determined that another overhead crane and the overhead crane where the visual detection device is located are on two intersecting lines, and the other overhead crane needs to turn at the intersection of the two lines, if it is determined that the overhead crane where the visual detection device is located will go straight through the intersection, then when the visual detection device on the overhead crane detects the other overhead crane that needs to turn at the intersection, the other overhead crane is made to stop running, and it is determined that the overhead crane where the visual detection device is located needs to enter the protection operation mode.

[0074] If it is determined that the overhead crane where the visual detection device is located will turn at the intersection, then when the visual detection device on the overhead crane detects the other overhead crane that needs to turn at the intersection, the overhead crane where the visual detection device is located is made to enter the protection operation mode, and the other overhead crane is made to turn through the intersection first.

[0075] After determining the protection operation mode, it is necessary to control the running speed of the overhead crane where the visual detection device is located according to the real-time distance between the other overhead crane and the overhead crane where the visual detection device is located. Therefore, it is necessary to determine the real-time distance between the other overhead crane and the overhead crane where the visual detection device is located. Specifically, based on the compensated rear parallax map, binocular visual images are calculated to obtain the real-time distance between the overhead crane where the visual detection device is located and the other overhead crane in front of it.

[0076] As shown in the Figure 3 appendix, the real-time distance is determined according to the following formula:

[0077] Z = (b * f) / d;

[0078] where Z is the real-time distance, b is the baseline length, f is the camera focal length, and d is the parallax, d = x L - x R , assuming that the optical axes of the two cameras are parallel and on the same horizontal line, then the abscissas of the projection points of the same target point in the left and right images are x L , x R , the specific determination method of the x L , x R is a known technology and will not be elaborated here.

[0079] In the protection operation mode, the moving speed of the overhead crane where the visual detection device is located is controlled according to the following process;

[0080] Determine whether the real-time distance is less than the speed reduction distance;

[0081] If not, make the overhead crane where the visual detection device is located travel at the normal moving speed and repeatedly determine whether the real-time distance reaches the speed reduction distance. The speed reduction distance can be set as needed. When the real-time distance is lower than the speed reduction distance, it indicates that the distance between the overhead crane where the visual detection device is located and another overhead crane in front of it is lower than the safe distance, and it is necessary to slow down to keep the distance between the overhead crane where the visual detection device is located and another overhead crane in front of it at the safe distance.

[0082] If so, make the overhead crane where the visual detection device is located run at a reduced speed and determine the change of the real-time distance;

[0083] If it is determined that the real-time distance is greater than the speed reduction distance, make the overhead crane where the visual detection device is located return to the normal moving speed;

[0084] If it is determined that the real-time distance is between the speed reduction distance and the stop distance, make the overhead crane where the visual detection device is located continue to travel at a reduced speed or maintain the current moving speed; the stop distance can be designed as needed and is less than the speed reduction distance.

[0085] If it is determined that the real-time distance is less than the stop distance, control the moving speed of the overhead crane where the visual detection device is located to be 0, and continue to judge whether the real-time distance exceeds the stop distance. If so, make the overhead crane where the visual detection device is located accelerate and judge the change of the real-time speed.

[0086] Embodiment 2

[0087] This embodiment discloses an overhead crane device, including a memory and a processor. A program executable by the processor is stored in the memory, and when the program is executed, it implements the overhead crane operation control method as described in any one of the above.

[0088] There are still many embodiments of the present invention. All technical solutions formed by equivalent transformation or equivalent substitution fall within the protection scope of the present invention.

Claims

1. The overhead crane operation control method is characterized in that: During the movement of an overhead crane, images collected by a vision detection device on the overhead crane are obtained in real time, and the traffic environment within a certain range in front of the overhead crane is determined based on the images collected by the vision detection device on the overhead crane obtained in real time and the information obtained from the dispatching system; Determine whether the overhead crane where the vision detection device is located needs to enter the protection operation mode at least according to the traffic environment within a certain range in front of the overhead crane; When entering the protection operation mode, control the moving speed of the vision detection device based on the real-time distance between the overhead crane where the vision detection device is located determined based on the images collected by the vision detection device obtained in real time and another overhead crane moving in the same direction in front of it; The traffic environment within a certain range in front of the overhead crane is one of the following situations, and the situations include no overhead crane, there is another overhead crane and it is moving in the same direction and maintaining a normal moving speed, there is another overhead crane and it is moving in the same direction and decelerating, there is another overhead crane and it is moving straight in the same direction and accelerating, there is another overhead crane stopped on its moving route, there is another overhead crane turning on its moving route; Determining the traffic environment within a certain range in front of the overhead crane based on the images collected by the vision detection device on the overhead crane obtained in real time and the information obtained from the dispatching system includes the following process: Input the images collected by the vision detection device obtained in real time into a trained target detection model after predetermined processing to determine the situation of overhead cranes within a certain range in front of the overhead crane, and the situation of overhead cranes is one of no overhead crane, there is another overhead crane and it is moving in the same direction, there is another overhead crane turning on its moving route; When it is determined that the situation of overhead cranes within a certain range in front of the overhead crane is no overhead crane or there is another overhead crane turning on its moving route, it is determined that the traffic environment within a certain range in front of the overhead crane is no overhead crane or there is another overhead crane turning on its moving route; When it is determined that the situation of overhead cranes within a certain range in front of the overhead crane is there is another overhead crane and it is moving in the same direction, determine whether the traffic environment within a certain range in front of the overhead crane is there is another overhead crane and it is moving in the same direction and decelerating, or there is another overhead crane moving straight in the same direction and accelerating, or there is another overhead crane stopped on its moving route, or there is another overhead crane moving straight in the same direction and maintaining a normal moving speed according to the speed situation of the other overhead crane obtained from the dispatching system.

2. The overhead crane operation control method according to claim 1, characterized in that: The predetermined processing of the images collected by the vision detection device obtained in real time includes: Perform image enhancement processing on the images collected by two cameras respectively to obtain an enhanced right view image and an enhanced left view image; Construct image pyramids corresponding to the enhanced right view image and the enhanced left view image respectively and obtain a left image feature map and a right image feature map based on the image pyramids; Perform preliminary disparity estimation on the lowest resolution layer based on the left image feature map and the right image feature map to obtain a disparity map; Obtain a compensated disparity map after compensating the disparity map.

3. The overhead crane operation control method according to claim 2, wherein: The image enhancement processing includes the following steps: First, perform histogram equalization on the image, then perform Laplacian operator enhancement, and finally perform Gamma correction.

4. The overhead crane operation control method according to claim 2, characterized in that: The compensation for the parallax map in sequence includes structured light compensation, temperature compensation, and vibration compensation.

5. The overhead crane operation control method according to claim 1, characterized in that: If it is determined that the traffic environment within a certain range in front of the overhead crane has another overhead crane traveling in the same direction and decelerating, or another overhead crane stopping on its moving route, or another overhead crane turning on its moving route, it is determined that the overhead crane where the visual detection device is located needs to enter the protection operation mode.

6. The overhead crane operation control method according to claim 1, characterized in that: The overhead crane where the visual detection device is located receives the running trajectory of the other overhead crane fed back by the dispatching system in real time. If it is determined that the traffic environment within a certain range in front of the overhead crane has another overhead crane traveling in the same direction and maintaining a normal moving speed, and the running trajectory of the other overhead crane fed back by the dispatching system is about to turn, it is determined that the overhead crane where the visual detection device is located needs to enter the protection operation mode.

7. The overhead crane operation control method according to claim 1, characterized in that: The overhead crane where the visual detection device is located receives the running trajectories of other overhead cranes fed back by the dispatching system in real time. If it is determined that another overhead crane and the overhead crane where the visual detection device is located are on two intersecting lines, and the other overhead crane needs to turn at the intersection of the two lines, if it is determined that the overhead crane where the visual detection device is located will go straight through the intersection, when the visual detection device on the overhead crane detects the other overhead crane that needs to turn at the intersection, stop the other overhead crane from traveling, and determine that the overhead crane where the visual detection device is located needs to enter the protection operation mode; If it is determined that the overhead crane where the visual detection device is located needs to turn at the intersection, when the visual detection device on the overhead crane detects the other overhead crane that needs to turn at the intersection, make the overhead crane where the visual detection device is located enter the protection operation mode, and make the other overhead crane turn through the intersection first.

8. The overhead crane operation control method according to claim 1, characterized in that: The real-time distance is determined according to the following formula: Z = (b * f) / d; where Z is the real-time distance, b is the baseline length, f is the camera focal length, and d is the parallax.

9. The overhead crane operation control method according to any one of claims 1-8, characterized in that: In the protection operation mode, control the moving speed of the overhead crane where the visual detection device is located according to the following process; Determine whether the real-time distance is less than the speed reduction distance; If not, make the overhead crane where the visual detection device is located travel at the normal moving speed and repeatedly determine whether the real-time distance reaches the speed reduction distance; If so, make the overhead crane where the visual detection device is located run at a reduced speed, and determine the change situation of the real-time distance; If it is determined that the real-time distance is greater than the speed reduction distance, make the overhead crane where the visual detection device is located resume the normal moving speed; If it is determined that the real-time distance is between the speed reduction distance and the stopping distance, make the overhead crane where the visual detection device is located continue to run at a reduced speed or maintain the current moving speed; If it is determined that the real-time distance is less than the stopping distance, control the moving speed of the overhead crane where the visual detection device is located to be 0, and continue to judge whether the real-time distance exceeds the stopping distance. If so, make the overhead crane where the visual detection device is located accelerate and judge the change situation of the real-time speed.

10. Overhead crane equipment, including a memory and a processor, wherein a program executable by the processor is stored in the memory, and is characterized in that: When the described program is executed, it implements the overhead crane operation control method according to any one of claims 1-9.

Images