Multi-door crane collaborative operation control method and system

Abstract

The present invention provides a multi-gantry crane collaborative operation control method and system, which relates to the technical field of multi-gantry crane collaborative operation. The method comprises the following steps: obtaining point cloud data inside a ship's cabin, constructing a digital elevation model, generating material distribution information, and extracting hatch boundary and hopper position information; dividing a grabbing window based on the digital elevation model and the material distribution information, calculating the material height, centripetality, and safety degree of each grabbing window, and comprehensively selecting a grabbing point; establishing a unified coordinate system for the multi-gantry cranes, integrating laser radar odometer information, IMU odometer information, and gantry crane motion structure constraints, and achieving high-precision positioning of the gantry crane's trunk beam through Kalman filtering; combining an artificial potential field and an A algorithm to plan a collision avoidance path for the multi-gantry cranes, and embedding timing information in the path points to ensure collaborative motion; and in the cabin clearing stage, obtaining the position information of the cabin clearing machine in real time, and dynamically adjusting the grabbing point to avoid collision.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-door crane cooperative operation, and in particular to a multi-door crane cooperative operation control method and system. Background Art

[0002] As a key node in the bulk resource supply chain, bulk cargo terminals handle the loading and unloading of strategic materials such as iron ore and coal. Their intelligence directly impacts national energy security and the resilience of the supply chain. As core loading and unloading equipment at bulk cargo terminals, the operational capacity of gantry cranes directly impacts the overall operational efficiency of the port.

[0003] However, traditional gantry crane operations rely primarily on visual observation and voice interaction by operators, making it difficult to achieve real-time and accurate perception of the motion status of multiple gantry cranes. This leads to delayed coordinated response and inaccurate collision avoidance decisions, seriously threatening operational safety. Furthermore, the human-machine interaction mechanism in complex operating environments has inherent flaws. Operators must synchronize the movements of two cranes in high-dust, high-noise conditions, resulting in attention overload and degraded operational accuracy. Furthermore, manual collaborative operations are limited by the cumulative effects of physiological fatigue, resulting in nonlinear attenuation of system stability with operating time, leading to frequent unplanned interruptions. These technical flaws have resulted in significant operational issues with existing gantry cranes, such as low efficiency and prominent safety hazards, severely hindering the intelligent upgrade process of bulk terminals. Summary of the Invention

[0004] The object of the present invention is to provide a method and system for controlling the collaborative operation of multiple gantry cranes, so as to solve the problems of low operating efficiency and prominent safety hazards in the operation of existing gantry cranes mentioned in the above background technology.

[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a method for controlling the collaborative operation of multiple gantry cranes, comprising the following steps: obtaining point cloud data inside the cabin, constructing a digital elevation model, generating material distribution information, and extracting hatch boundary and hopper position information; dividing the grabbing window based on the digital elevation model and the material distribution information, calculating the material height, centripetality, and safety degree of each grabbing window, and comprehensively selecting the grabbing point; establishing a unified coordinate system for multiple gantry cranes, installing an odometer on the trunk of the gantry crane, integrating the laser radar odometer information, the IMU odometer information, and the gantry crane motion structure constraints, and locating the position information of the trunk in the gantry crane through the Kalman filter algorithm; based on the selected grabbing point and the position information of the trunk, combining the artificial potential field and The algorithm plans collision avoidance paths for the multi-door cranes and embeds timing information in the path points to ensure coordinated movement between the multi-door cranes; it obtains the position information of the tank cleaning machine in real time and dynamically adjusts the gripping points to avoid collisions between the multi-door cranes and the tank cleaning machine.

[0006] Optionally, the steps of constructing a digital elevation model, generating material distribution information and extracting hatch boundary and hopper position information specifically include: projecting the point cloud data collected in the cabin onto the XOY plane, and constructing a plane triangulated mesh using a triangulated mesh generation algorithm; adding elevation information to the plane triangulated mesh, and generating a spatial irregular triangulated mesh to simulate the surface morphology of the material in the cabin; calculating the deformation of each spatial triangle relative to its plane projection, and identifying the surface feature area of ​​the material based on the change in the sine value of the triangle's internal angle; performing double circle feature analysis on the endpoints of each edge of the triangulated mesh, and extracting candidate boundary features through a circle center positioning algorithm with a preset radius; using an iterative maximum inscribed circle calculation method to determine the mid-axis point of the hatch boundary, and reconstructing the complete geometric boundary of the hatch and the hopper in combination with normal vector constraints.

[0007] The mid-axis point of the boundary is combined with the normal vector constraint to reconstruct the complete geometric boundaries of the hatch and hopper.

[0008] Optionally, the step of calculating the material height, centripetality and safety of each grabbing window to comprehensively select the grabbing point specifically includes: dividing the bottom plane into M×N cells, each cell has a side length of d and an average height of , forming a three-dimensional matrix to represent the distribution of the entire bottom material; dividing the grabbing window formed by the fully opened grabber into U×V grabbing sub-windows, and the grabbing sub-window located at the center of the grabbing window is the grabbing center point; calculating the average height of each grabbing sub-window and center point height , based on average height and center point height The ratio relationship of the material centripetal aggregation degree is established to quantify the centripetal degree; the safety degree is introduced Evaluate the height difference of the materials on both sides of the grasping window. When the center height is higher than both sides, the safety factor of the safety is 1, otherwise the safety factor of the safety is 0. Calculate the comprehensive evaluation coefficient K by dynamically adjusting the weights of centripetality and safety, and traverse all windows to select the grasping window with the largest K value as the optimal grasping point.

[0009] Optionally, the steps of establishing a unified coordinate system for multiple gantry cranes, installing an odometer on the trunk of the gantry crane, fusing the lidar odometer information, IMU odometer information and the gantry crane motion structure constraints, and locating the position information of the trunk in the gantry crane through the Kalman filter algorithm specifically include: calibrating the position information of several gantry cranes in the coordinate system to form a unified coordinate system for multiple gantry cranes; obtaining the lidar odometer information and IMU odometer information at the trunk of the gantry crane; obtaining the rotation angle of the gantry crane, the movement angle of the gantry crane arm and the gantry crane structure length to construct constraints based on the gantry crane motion structure; using Kalman filtering to fuse the lidar odometer information, IMU odometer information and the gantry crane motion structure constraints, thereby achieving high-precision positioning of the gantry crane trunk.

[0010] Optionally, the position information of the selected grasping point and the trunk bridge is combined with the artificial potential field and The algorithm plans collision avoidance paths for multiple cranes and trunks, and embeds timing information in the path points to ensure the coordinated movement of multiple cranes. The specific steps include: real-time data exchange between the cranes through TCP / IP communication; when it is detected that the distance between the trunks of two adjacent cranes is less than the set threshold, a repulsive potential field is introduced into the path planning. The size of the repulsive potential field is proportional to the distance between the positions of the two cranes' trunks, and the direction is the direction of the line connecting the positions of the two trunks; when it is detected that the distance between the trunks of two adjacent cranes is less than the minimum safe distance, a circular obstacle is generated with the end point of the trunk as the center, and the radius of the obstacle is proportional to the inverse square root of the minimum safe distance. At this time, the obstacle is generated. The algorithm performs detours; the motion paths of two adjacent gantry cranes are divided into several cycles of non-interfering actions, and timing information is embedded in each action. The actions of two adjacent gantry cranes in the same time period are different to ensure that interference is avoided during coordinated movement.

[0011] Optionally, the step of obtaining the position information of the tank cleaning machine in real time and dynamically adjusting the gripping point to avoid collision between the multi-door machine and the tank cleaning machine specifically includes: obtaining the positions of two reflective positioning plates of different sizes at the rear of the tank cleaning robot and the top of the cab, and generating a saturated collision space model for the left and right turns of the tank cleaning robot; at the initially selected gripping point A cylindrical collision avoidance zone is established around the bottom radius ,in is the grab radius, is the width of the tank cleaning machine, To provide more space for the grab bucket's shaking, the system uses a CropBox filter to crop the point cloud data within the cylindrical area, and then reselects alternative grab points from the remaining point clouds. The position of the cabin cleaning robot's reflective cursor plate is detected in real time, and the robot automatically switches to the alternative grab point when it enters the collision avoidance zone.

[0012] Optionally, the trunk beam is controlled by a double-layer PID control, specifically including: establishing a plane coordinate system with the gantry crane axis seat as the coordinate origin, and obtaining the actual position coordinates and the expected path point coordinates of the gantry crane arm end in real time; when performing high-level PID control, an angle error signal is generated by comparing the polar coordinate angle difference between the actual position and the expected position, and a radius error signal is generated by comparing the actual rotation radius with the expected rotation radius, and the composite error signal is processed by proportional-integral-differential operation to output a speed instruction including the target angular velocity of the rotating shaft and the target linear velocity of the pull rod; when performing low-level PID control, the actual angular velocity of the rotating shaft is obtained by the speed sensor, and the actual linear velocity of the pull rod is obtained by differentiating the displacement sensor; the difference between the actual speed and the target speed is processed by proportional-integral-differential operation to generate a motor drive current signal; and finally the drive signal is output to the rotary motor and the luffing mechanism gearbox respectively to realize the coordinated control of the gantry crane rotation and the arm extension.

[0013] On the other hand, the present invention also provides a multi-gantry crane collaborative operation control system, including: an acquisition module for acquiring point cloud data inside the cabin, building a digital elevation model, generating material distribution information and extracting hatch boundary and hopper position information; a grabbing point selection module for dividing the grabbing window based on the digital elevation model and the material distribution information, calculating the material height, centripetality and safety of each grabbing window, and comprehensively selecting the grabbing point; a positioning module for establishing a unified coordinate system for multiple gantry cranes, installing an odometer on the trunk of the gantry crane, integrating the laser radar odometer information, the IMU odometer information and the gantry crane motion structure constraint, and locating the position information of the trunk in the gantry crane through the Kalman filter algorithm; a collaborative module for combining the artificial potential field and the position information of the selected grabbing point and the trunk based on the selected The algorithm plans collision-avoidance paths for the movement of multi-door cranes and the trunk beam, and embeds timing information in the path points to ensure the coordinated movement of the multi-door cranes; the gripping point adjustment module is used to obtain the position information of the tank cleaning crane in real time and dynamically adjust the gripping points to avoid collisions between the multi-door cranes and the tank cleaning cranes.

[0014] On the other hand, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned multi-door machine collaborative operation control method when executing the computer program.

[0015] On the other hand, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the above-mentioned multi-door machine collaborative operation control method when executed by a processor.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] 1. This patented system achieves comprehensive awareness of the ship's operating environment through high-precision material distribution, hatch, and hopper identification technology. Through laser scanning and 3D modeling, the system accurately identifies material distribution and the locations of hatches and hoppers, ensuring precise positioning and operation of the gantry crane during unloading. This prevents operational errors caused by positioning errors and improves operational efficiency and safety.

[0018] 2. Regarding gripping point selection, the system dynamically selects the optimal gripping point by analyzing the Digital Elevation Model (DEM) and material distribution characteristics, ensuring maximum material yield per grab. Furthermore, combined with safety factor calculations, the system ensures that the grab bucket will not tip over due to material tilt during grabbing. By setting multiple gripping points and detecting the laser reflectors of the tank cleaning robot, the system effectively prevents collisions between the gantry crane and the tank cleaning robot, ensuring the safety and stability of the grabbing operation.

[0019] 3. In terms of door crane coordinated movement, this patent adopts a multi-door crane coordinated control strategy that takes into account timing information, and ensures the coordinated operation of the main and auxiliary door cranes through the TCP communication mechanism. The algorithm performs path planning to avoid collisions between door cranes and obstacles. The use of double-layer PID control allows the door crane to move smoothly according to the established instructions, further improving operation efficiency and system safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Schematic diagram of the process steps of the present invention.

[0021] Figure 2 It is a schematic diagram of the overall structure of the door machine of the present invention.

[0022] Figure 3 It is a schematic diagram of the overall structure of the tank cleaning machine of the present invention.

[0023] Figure 4 It is a schematic diagram of the collision space model structure of the tank cleaning machine of the present invention.

[0024] Figure 5 Schematic diagram of the digital elevation model of the present invention.

[0025] Figure 6 This is a schematic diagram of the effect of grabbing point selection in the present invention.

[0026] Figure 7 This is a schematic diagram of the motion structure constraints of the door crane of the present invention.

[0027] Figure 8 This is a schematic diagram of multi-door machine collaboration after the timing information is embedded in the present invention.

[0028] Figure 9 It is a schematic diagram of the double-layer PID control process of the present invention.

[0029] Figure 10 Schematic diagram of the system structure of the present invention.

[0030] In the figure: 1-camera, 2-laser radar, 3-wire sensor, 4-angle sensor, 5-reflective cursor positioning board, 10-acquisition module, 20-grasping point selection module, 30-positioning module, 40-cooperation module, 50-grasping point adjustment module. DETAILED DESCRIPTION

[0031] The following will provide a clear and complete description of the solutions of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] It will be understood by those skilled in the art that, unless expressly stated otherwise, the singular forms "a", "an", "said" and "the" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the specification of this application refers to the presence of features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, "connected" or "coupled" as used herein may include wireless connections or wireless couplings. The term "and / or" used herein includes all or any units and all combinations of one or more associated listed items.

[0034] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0035] It should be understood that the sequence numbers and sizes of the steps in this embodiment do not imply the order of execution. The order of execution of each process is determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of this application.

[0036] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0037] Please refer to Figures 1-9 The present invention provides a method for controlling the coordinated operation of multiple gantry cranes, comprising the following steps: obtaining point cloud data inside the cabin, constructing a digital elevation model, generating material distribution information, and extracting hatch boundary and hopper position information; dividing the grabbing windows based on the digital elevation model and the material distribution information, calculating the material height, centripetality, and safety degree of each grabbing window, and comprehensively selecting grabbing points; establishing a unified coordinate system for multiple gantry cranes, installing an odometer on the trunk beam of the gantry crane, integrating laser radar odometer information, IMU odometer information, and gantry crane motion structure constraints, and locating the position information of the trunk beam in the gantry crane through a Kalman filter algorithm; based on the selected grabbing points and the position information of the trunk beam, combining artificial potential field and The algorithm plans collision avoidance paths for the multi-door cranes and embeds timing information in the path points to ensure coordinated movement between the multi-door cranes; it obtains the position information of the tank cleaning machine in real time and dynamically adjusts the gripping points to avoid collisions between the multi-door cranes and the tank cleaning machine.

[0038] Specifically, this solution achieves fully automated operations by building a digital elevation model, employing multi-source information fusion positioning technology to improve spatial perception accuracy, and establishing a multi-machine collaborative mechanism through time-series path planning. A dynamic grip point adjustment strategy forms a closed-loop control system, effectively addressing the response delays and decision-making errors associated with traditional manual operations, significantly improving operational efficiency and effectively avoiding potential safety hazards.

[0039] In some embodiments, the steps of constructing a digital elevation model, generating material distribution information and extracting hatch boundary and hopper position information specifically include: projecting the point cloud data collected in the cabin onto the XOY plane, and constructing a plane triangulated mesh using a triangulated mesh generation algorithm; adding elevation information to the plane triangulated mesh, and generating a spatial irregular triangulated mesh to simulate the surface morphology of the material in the cabin; calculating the deformation of each spatial triangle relative to its plane projection, and identifying the surface feature area of ​​the material based on the change in the sine value of the triangle's internal angle; performing double circle feature analysis on the endpoints of each edge of the triangulated mesh, and extracting candidate boundary features through a circle center positioning algorithm with a preset radius; using an iterative maximum inscribed circle calculation method to determine the mid-axis point of the hatch boundary, and reconstructing the complete geometric boundary of the hatch and hopper in combination with normal vector constraints.

[0040] Specifically, the point cloud data is first projected onto the XOY plane, and then the Delaunay triangulated mesh of the plane is generated using the triangulated mesh generation algorithm. The generated plane irregular triangulated mesh is then added with elevation information to form an irregular triangulated mesh in space. The spatial irregular triangulated mesh is used to approximately simulate the actual situation of materials in the complex cabin.

[0041] ; Where: the deformation of the triangle It reflects the sum of the sine values ​​of the corresponding internal angle changes in the triangle. is a constant, is the interior angle of the triangle in the XoY plane, After adding elevation information to form spatial triangles, The corresponding interior angle.

[0042] The deformation of the triangle All edges e whose endpoints are and , find two radii The circle with the center being and :

[0043] Where: represents the length of the edge with endpoints p and q, is the x-coordinate of point p, is the x-coordinate of point q, is the y coordinate of point p, is the y coordinate of point q.

[0044] To determine the boundaries of the hatch , first we need to find the mid-axis point of the edge curve , and need to calculate The maximum inscribed circle of all attached point clouds p. For a given circle , k represents the number of iterations, N is the normal vector, the center of the circle for:

[0045] ; Find the nearest hatch surface point , and then test for the maximum value with a circle defined by: ; When the center axis point is found, the iteration process will stop. The properties of each hatch edge point include the coordinates of the center axis point ,radius , the index of surface points p and q, the normal vector Finally, by using the medial axis transform descriptor, the geometric boundaries of the hatch and hopper can be completely reconstructed.

[0046] This application utilizes triangle deformation feature analysis combined with a dual-circle feature positioning algorithm to overcome the technical bottleneck of complex cabin boundary extraction. Through spatial geometric constraints and iterative calculation of surface features, a high-fidelity three-dimensional representation of the cabin interior is constructed, providing a precise topological foundation for subsequent operational decisions and significantly improving the system's environmental adaptability in complex operating conditions.

[0047] In some embodiments, the step of calculating the material height, centripetality and safety of each grabbing window to comprehensively select the grabbing point specifically includes: dividing the bottom plane into M×N cells, each cell has a side length of d and an average height of , forming a three-dimensional matrix to represent the distribution of the entire bottom material; dividing the grabbing window formed by the fully opened grabber into U×V grabbing sub-windows, and the grabbing sub-window located at the center of the grabbing window is the grabbing center point; calculating the average height of each grabbing sub-window and center point height , based on average height and center point height The ratio relationship of the material centripetal aggregation degree is established to quantify the centripetal degree; the safety degree is introduced Evaluate the height difference of the materials on both sides of the grasping window. When the center height is higher than both sides, the safety factor of the safety is 1, otherwise the safety factor of the safety is 0. Calculate the comprehensive evaluation coefficient K by dynamically adjusting the weights of centripetality and safety, and traverse all windows to select the grasping window with the largest K value as the optimal grasping point.

[0048] Specifically, the capture window is divided, the average height of the central zone is calculated, and the mass distribution centripetality of each unit window is obtained. Figure 9 As shown, the entire bottom plane is divided into M×N cells, denoted as , where the side length of each cell is d and the average height is , and the entire area is represented as a three-dimensional matrix. When defining the material to be gripped, the gripper's fully opened gripping window is recorded as , each capture window includes U×V cells, and the cell at the center of the capture window is The grab center.

[0049] ;

[0050] Where: When determining the grabbing window, priority should be given to areas where the material height is high and concentrated. If the material distribution height in the center of the grabbing window is low and not on the "peak", the amount of material captured will be reduced, and the requirements of grabbing full bucket rate and high efficiency cannot be met. In order to ensure that the efficiency of each grab meets the requirements, this patent designs a grabbing window First, in order to conveniently express the material distribution in the grab window, Divided into multiple sub-windows , each subwindow contains Unit window The average height of the captured sub-window is:

[0051] ; The height of the grab center subwindow is , the centripetal model of the mass distribution of each grabbing sub-window The definition is as follows: ; Where: centripetal Indicates the degree to which matter in any grabbing window gathers toward the center, and its characteristic value The higher the height of the grabbing sub-window, the closer the material is to the center point of the window, thus generating a greater centripetal force. At the same time, the material distribution of the entire grabbing window is the average value of all sub-windows.

[0052] Capture Window Expectations : ;

[0053] It should be understood that the finer the division of the number of sub-windows, the more accurate the data expression. Considering the high variability of material distribution, it is necessary to reduce the tilt angle of the gripper when grabbing materials to reduce the possibility of accidents. The ideal distribution of materials under the gripping window is high in the middle and low around. The safety characteristic factor is defined as: ;in 、 、 and They are the average heights of the materials at the top, bottom, left and right sides of the grab window respectively. is the safety factor, satisfying the following conditions: ; Where: H represents the height difference tolerance value of the grabbing window. Once the height difference between the top and bottom or both sides of the grabbing window is greater than the tolerance value, the grabbing window is considered unsafe. The safety factor is 0, otherwise it is 1. Safety characteristic factor The larger it is, the safer the material handling process is.

[0054] Comprehensive evaluation coefficient K of the capture window: ; where a is the efficiency weight, b is the safety weight, and By adjusting the weights of a and b, the unloading needs of different types of unloading can be met. By traversing all windows in the cabin, the window with the highest K value is selected.

[0055] This application proposes a dual-dimensional evaluation system based on the centripetal model and dynamic safety assessment mechanism. This system enables refined analysis of material distribution characteristics through grid management and sub-window division. This solution establishes a mapping relationship between quantitative evaluation indicators and grasping strategies, fundamentally resolving the industry's challenge of optimizing both grasping efficiency and operational safety.

[0056] In some embodiments, the steps of establishing a unified coordinate system for multiple gantry cranes, installing an odometer on the trunk of the gantry crane, fusing the lidar odometer information, the IMU odometer information and the gantry crane motion structure constraints, and locating the position information of the trunk in the gantry crane through the Kalman filter algorithm specifically include: calibrating the position information of several gantry cranes in the coordinate system to form a unified coordinate system for multiple gantry cranes; obtaining the lidar odometer information and the IMU odometer information at the trunk of the gantry crane; obtaining the rotation angle of the gantry crane, the gantry crane arm motion angle and the gantry crane structure length to construct constraints based on the gantry crane motion structure; using Kalman filtering to fuse the lidar odometer information, the IMU odometer information and the gantry crane motion structure constraints, thereby achieving high-precision positioning of the gantry crane trunk.

[0057] Specifically, an angle sensor is installed at the gantry crane shaft to obtain the gantry crane rotation angle; a wire sensor is installed at the pull rod to obtain the specific depth of the pull rod; a laser radar is installed at the trunk beam, and the radar scans downward to sense the material distribution in the cabin, the hatch position and the hopper position; cameras are installed at the trunk beam and the control console, and the multi-gantry crane management subsystem obtains the streaming camera data screen for supervision.

[0058] The single door at the end of the door crane coordinate system in the x-axis direction is used as the starting door crane, and all door crane coordinate systems are calibrated to the coordinate system of this starting door crane. The starting door crane is door crane 1. Assume that there are m door cranes working together. Since the coordinate systems of each door crane are parallel and highly consistent, only the relative relationship between the x-axis and y-axis directions of each door crane coordinate system is considered. Define the translation matrix that calibrates the coordinate system of door crane A to the coordinate system of door crane 1: Where: Indicates the relative distance between door machine j and door machine 1 in the x-axis direction of the door machine 1 coordinate system. Indicates the relative distance between door machine j and door machine 1 in the y-axis direction of the door machine 1 coordinate system, Indicates that the door In the x-axis direction of the coordinate system, the door machine Distance door machine The relative distance, Indicates that the door In the y-axis direction of the coordinate system, the door machine Distance door machine The relative distance, assuming the coordinates of the gantry crane j coordinate system are , its coordinates in the gantry crane 1 coordinate system .

[0059] Through TCP / IP communication, the trunk beam coordinates and the axis center point coordinates of the gantry crane 1~m are sent to the multi-gantry crane management subsystem for subsequent path planning and control.

[0060] The positioning system is based on multi-information fusion, integrating the odometer, radar odometer, and IMU odometer calculated by the structure itself. The center position of the rotation axis is calibrated by the gantry crane's self-rotation.

[0061] A LiDAR and IMU are installed on the gantry crane's trunk to locate the crane. Before starting the positioning process, the IMU coordinate system must be calibrated to the LiDAR coordinate system. To ensure real-time positioning data, the IMU acquisition frequency must be consistent with the LiDAR during the preparation phase.

[0062] Positioning phase: Based on the existing LiDAR and IMU odometry, constraints based on the crane's extended structure were added to construct constraints based on the crane's kinematic structure. Kalman filtering was then used to fuse the three odometry information, resulting in a more accurate positioning of the crane's trunk.

[0063] Door machine rotation radius Rotation up / down angle of the door arm The relationship is:

[0064] ;

[0065] Where: In actual situations, the movement of the door crane's luffing mechanism is directly driven by the rack. The change map is changes.

[0066] and The relationship is: ; can be and The corresponding relationship is simplified as follows: In the movement of the gantry crane At this moment, when the rack length of the door crane luffing mechanism is When the differential of the rack length Differential of the overall radius length of the gantry crane's luffing mechanism The relationship is: Where: For the function The first derivative at .

[0067] It can be deduced that: ; Use the angle sensor at the door machine shaft to measure the line connecting the door machine arm and the shaft center and the door machine coordinate system The angle between the axes is The gantry crane coordinate system is established with the parallel shore as the x-axis, the vertical shore pointing to the coastline as the y-axis, and the vertical upward as the z-axis. By decomposing the gantry crane motion, we can get The coordinates of the trunk beam of the gantry crane in the gantry crane coordinate system at this moment for: ; First, the ICP matching method is used to calculate the inter-frame odometer of the laser radar data to obtain the pose On this basis, the radar coordinate system and the IMU coordinate system need to be calibrated to the gantry crane coordinate system. While keeping the extension length of the gantry crane's variable length mechanism unchanged, the rotating mechanism rotates 15 circles and records the trajectory point information of the gantry crane's trunk beam. .

[0068] Rotation center coordinates of the door machine's rotation mechanism The rotation radius during calibration Use the following formula to calculate:

[0069] ; For trajectory points The accumulated average.

[0070] Calculated by the following formula:

[0071] ; and The residual between Calculated by the following formula: ; You can set it yourself here The acceptable threshold of If this threshold is exceeded, it is considered that the calibration of the door crane geometric parameters or the internal and external parameters of the laser radar and IMU in the system has failed, and re-measurement and calibration are required. In the preparation stage, the rotation center of the rotating mechanism must also be calibrated. Assume that the coordinate system position relationship is , where R represents the rotation matrix and t represents the translation matrix. Through field measurement, calibrate the IMU coordinate system to the radar coordinate system , calibrate the radar coordinate system to the gantry crane coordinate system After calibrating the positioning results of the gantry crane's trunk beam to the gantry crane coordinate system, the Kalman filter is used to fuse the three positioning results to obtain .

[0072] This application integrates a triple-sensing architecture consisting of lidar, IMU, and kinematic constraints for gantry cranes to construct a positioning system capable of compensating for mechanical characteristics. Kalman filtering is used to achieve spatiotemporal alignment of heterogeneous multi-source data, establishing a unified coordinate system to break down information silos and achieve a positioning method that combines millimeter-level accuracy with high robustness.

[0073] In some embodiments, the position information of the selected grasping points and the trunk bridge is combined with the artificial potential field and The algorithm plans collision avoidance paths for multiple cranes and trunks, and embeds timing information in the path points to ensure the coordinated movement of multiple cranes. The specific steps include: real-time data exchange between the cranes through TCP / IP communication; when it is detected that the distance between the trunks of two adjacent cranes is less than the set threshold, a repulsive potential field is introduced into the path planning. The size of the repulsive potential field is proportional to the distance between the positions of the two cranes' trunks, and the direction is the direction of the line connecting the positions of the two trunks; when it is detected that the distance between the trunks of two adjacent cranes is less than the minimum safe distance, a circular obstacle is generated with the end point of the trunk as the center, and the radius of the obstacle is proportional to the inverse square root of the minimum safe distance. At this time, the obstacle is generated. The algorithm performs detours; the motion paths of two adjacent gantry cranes are divided into several cycles of non-interfering actions, and timing information is embedded in each action. The actions of two adjacent gantry cranes in the same time period are different to ensure that interference is avoided during coordinated movement.

[0074] Specifically, for bulk carrier unloading operations in the port area, multiple gantry cranes work simultaneously in the berth, and a time sequence based on artificial potential field is designed. The single-door machine subsystem sends the position of the axle seat and the trunk beam to the multi-door machine management subsystem in real time through TCP / IP communication, and the multi-door machine management subsystem simultaneously performs motion planning based on the time sequence of the artificial potential field. After the algorithm is implemented, the path planning points with timing information are sent to each single-door machine subsystem via TCP / IP communication. During the algorithm's operation, if the distance between the two trunks is found to be too small, a circular obstacle will be set up with the end points of the two trunks as the center, and the two trunks will avoid it.

[0075] Tradition The algorithm will choose to detour when encountering obstacles, which is suitable for path planning for a single moving object. This patent is used in multi-door machines. In order to enable the multi-door machines to cooperate, it is necessary to enable the multi-door machines to add timing information to achieve collaborative path planning for multiple door machines.

[0076] When two adjacent gantry cranes A and B move between the hatch and the hopper, there is a risk of collision. The algorithm plans the trajectory of multiple aircraft, which operate in a grid map. When the algorithm is used, the grid points of the planned path will be accompanied by timing information, and the operating speed of multiple door machines will be assumed to be consistent, that is, the timing information of the i-th path planning grid of multiple door machines will be consistent, so as to realize multi-door and simultaneous path planning.

[0077] When the distance reaches the threshold, repulsion is added to the multi-door machine and added to the path planning In the algorithm, multiple door cranes are guided to avoid each other during movement and plan the shortest path while maintaining safety. Take door crane A as an example: Where: It is the evaluation function of node n, which represents the estimated cost from the starting node to node n and then from node n to the target node. is the actual cost from the starting node to node n. is the estimated cost from node n to the target node, such as the Euclidean distance. : repulsive force in the gravitational potential field. When the distance between the gripper and the robot is less than the threshold When the multi-door aircraft has repulsive force during path planning . Where: is the gain coefficient of the gravitational potential field, is the position of the trunk of the door crane A in the same time sequence, is the position of the trunk of the door crane B in the same time sequence; It is the unit vector from the trunk position of gantry crane B to the trunk position of gantry crane A, indicating the direction. It is obtained from the following formula: Where: is the position vector of the trunk of the gantry crane A at the same time sequence, is the position vector of the trunk of gantry crane B in the same time sequence.

[0078] The main tasks of the gantry crane are: 1. Grab materials from the cabin, 2. Move them from the hatch to the receiving hopper, 3. Place the materials in the receiving hopper, and 4. Move them from the receiving hopper to the hatch. The main task of the cabin cleaning robot is to shovel materials into a pile.

[0079] Taking two adjacent gantry cranes, A and B, as an example, the time intervals for the multiple gantry cranes' grabbing process and the interspersed material production actions of the cleaning robot are recorded as t1 to t5. The four time periods t1 to t4 together constitute a complete cycle of the system.

[0080] In order to enable the two cranes to operate in coordination, the TCP communication mechanism is used. The A and B cranes send the identifiers of the completed task nodes within the system to achieve coordinated movement. The A crane grabs the material from the hatch as the first node and creates the identifier F. Specifically: Default and The value is 0, which is sent when the i-th node A gate machine task is completed. =1, B gate machine mission completed and sent =1. When the two door machines receive and When both are 1, the next node can be run, otherwise the waiting strategy is adopted. and When both are 1, the cleaning robot starts shoveling materials. When i is an even number and and When both are 1, the tank cleaning robot stops shoveling materials.

[0081] When the gantry cranes are running between the hatch and the hopper, there is a risk of collision between the trunks of the adjacent gantry cranes. When , a circular obstacle is automatically generated between the two trunks of the machine. The circular coordinates of the circular obstacle are the end points of the two trunks, and the radius is When it is detected that the distance between the two trunks is greater than , delete the circular obstacle.

[0082] Specifically, this application combines artificial potential field theory with timing constraints and Combining these algorithms, a system for predicting and resolving multi-machine motion conflicts is established through the dynamic generation of repulsive fields and obstacle modeling. This solution achieves a qualitative shift from passive obstacle avoidance to active collaboration, providing a universal solution for multi-machine collaboration in intensive operation scenarios.

[0083] In some embodiments, the steps of obtaining the position information of the tank cleaning machine in real time and dynamically adjusting the gripping point to avoid collision between the multi-door crane and the tank cleaning machine specifically include: obtaining the positions of two reflective positioning plates of different sizes at the rear of the tank cleaning robot and the top of the cab, and generating a saturated collision space model for the left and right turns of the tank cleaning robot; at the initially selected gripping point A cylindrical collision avoidance zone is established around the bottom radius ,in is the grab radius, is the width of the tank cleaning machine, To provide more space for the grab bucket's shaking, the system uses a CropBox filter to crop the point cloud data within the cylindrical area, and then reselects alternative grab points from the remaining point clouds. The position of the cabin cleaning robot's reflective cursor plate is detected in real time, and the robot automatically switches to the alternative grab point when it enters the collision avoidance zone.

[0084] Specifically, during the cleaning phase, when the grab bucket is performing the grabbing operation, there is a cleaning robot inside the cabin responsible for shoveling the remaining bulk materials into a pile. At this time, it is necessary to ensure that the gantry crane does not collide with the cleaning robot when grabbing. The saturation selection method of grabbing points is used. The specific method process is as follows: After the gantry crane's trunk beam radar scans the material, it selects a suitable grabbing point, which is called grabbing point 1. Based on the original point cloud, the CropBox filter in PCL is used to crop the points with the material. The cylindrical space point cloud is a circle with the center of the bottom circle, R as the radius of the bottom circle, and h = 5m. In the formula, r is calculated by the following formula: Where: Indicates the grab radius range of the gantry crane grab bucket. Indicates the width of the cabin cleaning machine. This represents the excess space for the grab bucket's sway, used to account for the sway that occurs when the gantry crane grab bucket is lifted. A second suitable grab point is selected from the cropped point cloud, referred to as grab point 2. During the gantry crane's grabbing phase, the radar scans the position of the cargo clearing machine on the crane's trunk to determine whether the distance between the radar scanned cargo clearing machine position and the first grab point is greater than R. If so, the first grab point is selected; if less than R, the second grab point is selected.

[0085] This application builds a motion model for the tank-clearing robot based on a 3D sensing solution using a reflective fixed plate. By dynamically adjusting the cylindrical collision avoidance zone, it forms a safety protection system for human-robot collaborative operations. Real-time point cloud cropping and a rapid decision-making mechanism for alternative points ensure the system's continued reliable operation during the tank-clearing phase.

[0086] In some embodiments, the trunk beam is controlled by a double-layer PID control, specifically including: establishing a plane coordinate system with the gantry crane axis seat as the coordinate origin, and obtaining the actual position coordinates and the expected path point coordinates of the gantry crane arm end in real time; when performing high-level PID control, an angle error signal is generated by comparing the polar coordinate angle difference between the actual position and the expected position, and a radius error signal is generated by comparing the actual rotation radius with the expected rotation radius, and the composite error signal is processed by proportional-integral-differential operation to output a speed instruction including the target angular velocity of the rotating shaft and the target linear velocity of the pull rod; when performing low-level PID control, the actual angular velocity of the rotating shaft is obtained by the speed sensor, and the actual linear velocity of the pull rod is obtained by differentiating the displacement sensor; the difference between the actual speed and the target speed is processed by proportional-integral-differential operation to generate a motor drive current signal; and finally, the drive signal is output to the rotary motor and the luffing mechanism gearbox respectively to realize the coordinated control of the gantry crane rotation and the arm extension.

[0087] Specifically, a two-layer PID control system is used for the movement of the door operator. The high-level PID control calculates the target speed based on the error between the target position and the actual position time. The low-level PID control outputs the drive current signal of the control motor based on the error between the target speed output by the high-level PID control and the actual speed, thereby adjusting the speed.

[0088] High-level PID control: take the door machine axis seat as the coordinate origin , assuming the actual path point of the gantry crane is According to the above positioning results, the expected path point of the gantry crane is From the above The algorithm is derived.

[0089] The error between the expected angle and the current angle for:

[0090] ; According to the distance formula Can be found separately 、 、 The actual and expected length of the door crane rotation radius 、 and 、 Same. Door machine rotation radius The error between the expected value and the current value for: ; Use PID closed-loop control method to control the position of the gantry crane. The input signal at time T is , which indicates the error between the expected value and the current value. Door machine speed By the door machine shaft speed and the extension speed of the pull rod composition . High-level PID controller output value for , calculated by the following formula: Where: is the proportional gain, which is the adjustment parameter; is the integral gain, which is the adjustment parameter; is the differential gain, which is the adjustment parameter; For the current time.

[0091] Low-level PID control: door machine speed By the door machine shaft speed Composed of the extension and retraction speed of the pull rod . Actual door crane speed for , It is directly obtained from the speed sensor at the door machine shaft. The output value of the wire sensor at the differential pull rod is obtained, and the high-level PID control output speed at time t is for The input value of the low-level PID control for , output value for It is calculated by the following formula:

[0092] Where: is the proportional gain, which is the adjustment parameter; is the integral gain, which is the adjustment parameter; is the differential gain, which is the adjustment parameter; Error = set value SP - feedback value PV; The double-layer PID control method outputs corresponding control signals to the slewing motor at the gantry crane arm shaft and the gearbox of the luffing mechanism, controlling the gantry crane's rotation and the extension and retraction of the boom.

[0093] This application's dual-layer PID control architecture overcomes the trajectory tracking bottleneck under high-inertia loads through kinematic decoupling and a composite control strategy. The coordinated optimization of the position and velocity loops forms a cascade control system, significantly improving the system's control quality and energy efficiency under variable loads and high-disturbance conditions.

[0094] Please refer to Figure 10On the other hand, the present invention also provides a multi-gantry crane collaborative operation control system, including: an acquisition module for acquiring point cloud data in the cabin, building a digital elevation model, generating material distribution information and extracting hatch boundary and hopper position information; a grabbing point selection module for dividing the grabbing window based on the digital elevation model and the material distribution information, calculating the material height, centripetality and safety degree of each grabbing window, and comprehensively selecting the grabbing point; a positioning module for establishing a unified coordinate system for multiple gantry cranes, installing an odometer on the trunk beam of the gantry crane, integrating the laser radar odometer information, the IMU odometer information and the gantry crane motion structure constraint, and locating the position information of the trunk beam in the gantry crane through the Kalman filter algorithm; a collaborative module for combining the artificial potential field and the selected grabbing point and the position information of the trunk beam based on the selected grabbing point The algorithm plans collision-avoidance paths for the movement of multi-door cranes and the trunk beam, and embeds timing information in the path points to ensure the coordinated movement of the multi-door cranes; the gripping point adjustment module is used to obtain the position information of the tank cleaning crane in real time and dynamically adjust the gripping points to avoid collisions between the multi-door cranes and the tank cleaning cranes.

[0095] On the other hand, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned multi-door machine collaborative operation control method when executing the computer program.

[0096] On the other hand, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the above-mentioned multi-door machine collaborative operation control method when executed by a processor.

[0097] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing related hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the above-described method embodiments. In the embodiments provided by the present invention, any reference to memory, storage, database, or other media can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0098] The above are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's description and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for controlling the coordinated operation of multiple cranes, characterized in that the steps include: Obtain point cloud data inside the cabin, build a digital elevation model, generate material distribution information, and extract hatch boundary and hopper location information; Dividing the grabbing windows based on the digital elevation model and the material distribution information, calculating the material height, centripetality and safety of each grabbing window, and comprehensively selecting the grabbing points; Establish a unified coordinate system for multiple portal cranes, install an odometer on the trunk of the portal crane, integrate the lidar odometer information, the IMU odometer information, and the portal crane motion structure constraints, and use the Kalman filter algorithm to locate the position information of the trunk of the portal crane; Based on the selected grasping points and the position information of the trunk, combined with the artificial potential field and The algorithm plans collision avoidance paths for multiple aircraft and embeds timing information in the path points to ensure coordinated movement among the aircraft. The position information of the tank cleaning machine is obtained in real time, and the grabbing point is dynamically adjusted to avoid collision between the multi-door crane and the tank cleaning machine.

2. The multi-door crane collaborative operation control method according to claim 1, characterized in that: The steps of constructing a digital elevation model, generating material distribution information, and extracting hatch boundary and hopper location information specifically include: Project the point cloud data collected in the cabin onto the XOY plane and construct a plane triangulated mesh using a triangulated mesh generation algorithm; Adding elevation information to the plane triangulated mesh to generate a spatial irregular triangulated mesh to simulate the surface morphology of materials in the cabin; Calculate the deformation of each spatial triangle relative to its plane projection, and identify the surface feature area of ​​the material based on the change in the sine value of the triangle's internal angle; Perform double circle feature analysis on the endpoints of each edge of the triangulated network and extract candidate boundary features through the circle center positioning algorithm with a preset radius; The iterative maximum inscribed circle calculation method is used to determine the mid-axis point of the hatch boundary, and the complete geometric boundaries of the hatch and hopper are reconstructed in combination with normal vector constraints.

3. The multi-door crane collaborative operation control method according to claim 1, characterized in that: The step of calculating the material height, centripetality and safety of each grabbing window to comprehensively select the grabbing point specifically includes: Divide the bottom plane into M×N cells, each with a side length of d and an average height of , forming a three-dimensional matrix to represent the distribution of the entire bilge material; Divide the grabbing window formed by fully opening the grabber into U×V grabbing sub-windows, and the grabbing sub-window located at the center of the grabbing window is the grabbing center point; Calculate the average height of each of the captured sub-windows and center point height , based on average height and center point height The material centripetal aggregation evaluation model is established based on the ratio relationship to quantify the centripetal degree; The safety factor SGuv is introduced to evaluate the height difference of materials on both sides of the gripping window. When the center height is higher than both sides, the safety factor of the safety factor is 1, otherwise the safety factor of the safety factor is 0. The comprehensive evaluation coefficient K is calculated by dynamically adjusting the weights of centripetality and safety, and the grabbing window with the largest K value is selected as the optimal grabbing point through traversing all windows.

4. The multi-door crane collaborative operation control method according to claim 1, characterized in that: The steps of establishing a unified coordinate system for multiple portal cranes, installing an odometer on the trunk of the portal crane, integrating laser radar odometer information, IMU odometer information, and portal crane motion structure constraints, and locating the position information of the trunk of the portal crane using a Kalman filter algorithm specifically include: Calibrate the position information of several gantry cranes in the coordinate system to form a unified coordinate system for multiple gantry cranes; Obtain the laser radar odometer information and IMU odometer information at the trunk of the gantry crane; Obtain the gantry crane's rotation angle, gantry crane arm motion angle, and gantry crane structure length to construct constraints based on the gantry crane's motion structure; Kalman filtering is used to fuse the lidar odometer information, IMU odometer information and gantry crane motion structure constraints to achieve high-precision positioning of the gantry crane trunk.

5. The multi-door crane collaborative operation control method according to claim 1, characterized in that: The method is based on the selected grasping point and the position information of the trunk, combined with the artificial potential field and The algorithm plans collision-avoidance paths for the multi-door aircraft and the trunk beam, and embeds timing information into the path points to ensure coordinated movement between the multi-door aircraft. The specific steps include: Real-time data exchange between door machines via TCP / IP communication; When it is detected that the distance between the trunks of two adjacent telescopes is less than the set threshold, a repulsive potential field is introduced into the path planning. The magnitude of the repulsive potential field is proportional to the distance between the two telescopes’ trunk positions, and its direction is the direction of the line connecting the two trunk positions. When it is detected that the distance between the trunks of two adjacent cranes is less than the minimum safe distance, a circular obstacle is generated with the end point of the trunk as the center. The radius of the obstacle is proportional to the inverse of the square root of the minimum safe distance. The algorithm performs a detour; The motion paths of two adjacent gantry cranes are divided into several cycles of non-interfering actions, and timing information is embedded in each action. The actions of two adjacent gantry cranes in the same time period are different to ensure that interference is avoided during coordinated movement.

6. The multi-door crane collaborative operation control method according to claim 1, characterized in that: The steps of obtaining the position information of the tank cleaning machine in real time and dynamically adjusting the gripping point to avoid collision between the multi-door crane and the tank cleaning machine specifically include: Obtain the positions of two reflective marking plates of different sizes at the rear of the cleaning robot and on top of the cab, and generate a saturated collision space model for the cleaning robot's front end turning left and right. At the initially selected grab point A cylindrical collision avoidance zone is established around the bottom radius ,in is the grab radius, is the width of the tank cleaning machine, It is used to indicate the shaking of the grab bucket when it is hoisted. Use the CropBox filter to crop the point cloud data within the cylindrical collision avoidance area, and reselect alternative grasping points in the remaining point cloud; The position of the reflective cursor plate of the cleaning robot is detected in real time, and the robot automatically switches to the backup grabbing point when it enters the collision avoidance area.

7. The multi-door crane collaborative operation control method according to claim 1, characterized in that: The trunk bridge is controlled by a double-layer PID, specifically including: Establish a plane coordinate system with the gantry crane axis seat as the coordinate origin, and obtain the actual position coordinates of the gantry crane boom end and the coordinates of the expected path point in real time; When performing high-level PID control, an angle error signal is generated by comparing the polar coordinate angle difference between the actual position and the desired position, and a radius error signal is generated by comparing the actual rotation radius and the desired rotation radius. The composite error signal is processed using proportional-integral-differential operations to output a speed command containing the target angular velocity of the shaft and the target linear velocity of the pull rod. During the low-level PID control, the actual shaft angular velocity is obtained through the speed sensor, and the actual pull rod linear velocity is obtained through the displacement sensor differentiation. The difference between the actual speed and the target speed is processed using proportional-integral-differential operations to generate the motor drive current signal. Finally, the drive signals are output to the slewing motor and the luffing mechanism gearbox respectively, realizing the coordinated control of the gantry crane rotation and the boom extension and retraction.

8. A multi-door crane collaborative operation control system, characterized in that: include: The acquisition module is used to obtain point cloud data inside the cabin, build a digital elevation model, generate material distribution information, and extract hatch boundary and hopper location information; A grabbing point selection module is used to divide the grabbing windows based on the digital elevation model and the material distribution information, calculate the material height, centripetality and safety degree of each grabbing window, and comprehensively select the grabbing points; The positioning module is used to establish a unified coordinate system for multiple portal cranes. An odometer is installed on the trunk of the portal crane. The module integrates the laser radar odometer information, the IMU odometer information, and the portal crane's motion structure constraints to locate the position of the trunk of the portal crane using the Kalman filter algorithm. Collaborative module, used to combine artificial potential field and The algorithm plans collision-avoidance paths for the multi-door aircraft and the trunk beam, and embeds timing information in the path points to ensure coordinated movement between the multi-door aircraft. The grabbing point adjustment module is used to obtain the position information of the tank cleaning machine in real time and dynamically adjust the grabbing point to avoid collision between the multi-door crane and the tank cleaning machine.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the multi-door machine collaborative operation control method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the multi-door machine collaborative operation control method described in any one of claims 1 to 7 are implemented.

Images