A fully automatic control system for a gantry-type crusher

Abstract

This invention discloses a fully automatic control system for a gantry crusher, relating to the field of bulk material handling technology. It aims to solve the problems of low efficiency, poor accuracy, significant safety hazards, and inability to adaptively iterate when manually or semi-automatically handling clogging of ore bins. The system integrates dual sensing devices: LiDAR and a 3D camera. The LiDAR scans to acquire 3D point clouds, while the 3D camera captures 3D video images and combines them with AI video analysis algorithms to achieve multi-source data fusion processing. After extracting the 3D spatial coordinates of the clogging material through target identification, these coordinates are converted to the robotic arm's base coordinate system. This allows for the planning of a collision-free motion trajectory for the breaker hammer and optimization of the impact sequence, controlling the robotic arm to automatically execute crushing operations sequentially. The system also integrates safety interlocks, iterative crushing cycles, and manual intervention modes. This invention significantly reduces reliance on manual labor, improves operational continuity, intelligence, and safety reliability, and is adaptable to different specifications of gantry crushers, making it widely applicable.

Description

Technical Field

[0001] This invention belongs to the field of bulk material processing technology, specifically, it relates to a fully automatic control system for a gantry-type crusher. Background Technology

[0002] In mining and material handling processes, the grating in the ore bin is a key component ensuring the normal operation of subsequent equipment. Its main function is to intercept large pieces of material, preventing them from entering the ore bin and clogging conveying equipment or damaging related machinery. When large pieces of material accumulate and block the top of the grating, they must be crushed in a timely manner; otherwise, it will affect the normal operation of the entire material conveying system, causing production stoppages and resulting in significant economic losses.

[0003] Currently, the crushing of materials clogging the mine's grid mainly relies on manual operation. The specific method involves operators assessing the blockage at the grid through on-site observation or monitoring, then manually maneuvering the gantry crusher's robotic arm and hammers to the location of the blockage and execute the crushing action. However, this manual operation method has significant technical limitations and fails to meet the demands of modern mines for efficient, safe, and intelligent production, as detailed below: 1. Low operational efficiency and poor continuity: The entire process of manually identifying blocked materials, judging material location, and controlling equipment movement and crushing relies on human experience. The operation is cumbersome and it is difficult to plan the optimal striking sequence for multiple blocked materials. Repeated striking or ineffective movement is likely to occur, resulting in long processing time for each blockage and failure to quickly restore the grid to its smooth state, which in turn affects the production continuity of the entire material conveying system. Furthermore, the grid's smoothness cannot be quickly checked after manual crushing, which can easily lead to secondary blockages and further reduce operational efficiency.

[0004] 2. Poor positioning accuracy and unstable crushing effect: Manual operation can only judge the approximate position of the material by vision, and cannot accurately obtain the three-dimensional spatial coordinates of the material. This results in a large deviation in the position of the breaker hammer, which not only increases the number of additional blows and prolongs the operation time, but may also damage the grid. At the same time, the operation effect is highly dependent on the experience of the operator. Different operators have different skill levels, making it difficult to guarantee the stability of crushing quality.

[0005] 3. Low operational safety and high labor intensity: The mining area has high dust levels, poor visibility, and materials are prone to secondary collapse. Whether operating on-site or remotely blindly, operators face high personal safety risks. Moreover, manual operation requires full focus on material identification and trajectory control. Long-term operation can easily lead to fatigue, which not only increases the probability of operational errors, but may also lead to the inability to avoid collisions between the robotic arm and surrounding equipment due to negligence, further exacerbating the risk of equipment damage.

[0006] To address the three core technical pain points of efficiency, accuracy, and safety in manual operations, there is an urgent need for a control system that can automatically identify, accurately locate, plan the optimal trajectory, and automatically crush blockage materials. This system would completely solve the drawbacks of manual operation, improve the automation and intelligence level of mine bin grid blockage crushing operations, and ensure production continuity and operational safety. Summary of the Invention

[0007] In order to overcome the above-mentioned technical problems, the purpose of this invention is to provide a fully automatic control system for a gantry-type crusher.

[0008] The objective of this invention can be achieved through the following technical solutions: A fully automatic control system for a gantry-type crusher includes the following steps: S1. Perception and Positioning: A lidar and a 3D camera are fixedly installed on the crossbeam of the gantry truss to perform 3D scanning and video acquisition of the grid area of ​​the mine below, and to obtain the original point cloud data and 3D video image data of the grid area in real time. Among them, the 3D video image data acquired by the 3D camera is used for AI video analysis, which complements the original point cloud data acquired by the lidar to improve the accuracy of perception. S2. Target Recognition and Extraction: The raw point cloud data and the three-dimensional video image data collected by the 3D camera are fused and processed. The AI ​​video analysis algorithm is used to segment and identify the material targets to be crushed that are blocked above the grid, and to calculate the three-dimensional spatial coordinates of each material target in the lidar coordinate system or the 3D camera coordinate system. S3. Coordinate Space Unification: Based on the preset fixed coordinate transformation relationship between the gantry truss and the lidar and 3D camera, the three-dimensional spatial coordinates of each material target are uniformly transformed to the base coordinate system of the crusher robotic arm, generating the corresponding crushing and impact target point coordinates. S4. Motion Planning and Sequence Generation: Based on the coordinates of the target point of the breaker, a collision-free motion trajectory is planned for the breaker, and an ordered strike sequence is generated by combining efficiency or distance principles. S5. Automatic execution: Controls the gantry truss crusher to drive the robotic arm and breaker hammer, and moves them to each target point in sequence according to the above motion trajectory and impact sequence, and performs automatic crushing actions.

[0009] As a preferred embodiment of the present invention, step S2 specifically includes: S21. Filter and denoise the original point cloud data to remove discrete noise points; perform median filtering and rectangle correction on the 3D video image data acquired by the 3D camera to eliminate redundant information caused by imaging distortion and environmental interference, and ensure the accuracy of image data. S22. The random sampling consistency plane fitting algorithm is adopted to fit the mathematical model of the grid plane from the denoised point cloud and the corrected 3D video image data respectively. The accuracy of the plane model is improved by cross-validation of multi-source data. S23. Point cloud and image pixels that are more than a preset height threshold away from the mathematical model of the grid plane are taken as the data set to be processed. The set is segmented by a density-based spatial clustering algorithm. Combined with the grid blockage identification model in AI video real-time analysis, each connected data cluster is identified as an independent material target. S24. Calculate the physical volume and three-dimensional centroid of the data cluster corresponding to each material target. Filter effective blockage targets through the reverse reasoning model of the break point. Determine the centroid coordinates of the material targets whose volume is greater than the preset blockage volume threshold as the three-dimensional spatial coordinates of the material target in the corresponding coordinate system.

[0010] As a preferred embodiment of the present invention, step S4, which involves planning a collision-free motion trajectory for the hydraulic breaker, specifically includes: S41. For the coordinates of each breaking and impact target point, and in combination with the attitude requirements of the end effector of the hydraulic breaker, the target angles of each joint are solved using the inverse kinematics model of the robotic arm. S42. Perform trajectory interpolation in joint space or Cartesian space. During the planning process, use simulation collision detection to ensure that the breaker rod and the robotic arm body maintain a safe distance from the grid and the surrounding environment.

[0011] As a preferred embodiment of the present invention, before step S1, the method further includes: S0. Safety Interlock: The active area of ​​the gantry crusher is monitored in real time by a sensor network deployed in the work area; when personnel or mobile equipment are detected to have entered the preset emergency stop area, an emergency stop command is immediately sent to the core controller of the crusher; the core controller responds to the emergency stop command, interrupts any automatic crushing operation that is being performed or about to be performed, and controls the breaker hammer to be raised to a safe stopping position.

[0012] As a preferred embodiment of the present invention, after completing one round of impact in step S5, an iterative breaking step is further included: Return to step S1 and scan the grid area again, acquire video, and identify the target using LiDAR and 3D camera; If a valid blockage target is identified again, steps S2 to S5 are repeated for the next round of crushing until the grid area is clear or the maximum number of iterations set by the system is reached.

[0013] As a preferred embodiment of the present invention, a manual intervention mode is also included. When switching to this mode: It receives direct motion commands from remote operating stations or local operating panels, and parses these commands into real-time control quantities for the servo motors of each joint of the robotic arm and the telescopic cylinder of the breaker, thereby realizing continuous, adjustable-speed manual control of the breaker's position and movement. Meanwhile, in manual intervention mode, all key status parameters of the crusher, alarm information, and real-time video footage from the 3D camera are fed back to the operation interface.

[0014] As a preferred embodiment of the present invention, in step S5, when controlling the breaker hammer to perform automatic crushing operations: The extension length of the chisel is detected in real time by a high-precision displacement sensor installed on the intelligent telescopic cylinder of the hydraulic breaker. The detected actual length value is compared with the theoretical safe length value calculated based on the grid plane and target point coordinates to form a closed-loop feedback; By dynamically adjusting the control signal of the intelligent telescopic cylinder, a preset dynamic working gap is maintained between the end of the chisel and the grid surface during the actual crushing process.

[0015] As a preferred embodiment of the present invention, the specific process of rectangular correction is as follows: Median filtering is used to denoise the 3D video image data acquired by the 3D camera; Extract different color channels (BGR) and detect rectangles separately; Canny edge detection or multi-threshold binarization is applied to each color channel; Use the findContours function to find image contours; The approxPolyDP function is used to remove minor undulations in the polygon outline; Select quadrilaterals with larger areas and convex shapes; Determine if the cosine of the angle between any two adjacent lines of the quadrilateral is less than 0.3. If so, the quadrilateral is determined to be the target rectangle, and the correction is completed.

[0016] As a preferred technical solution of the present invention, the AI ​​video analysis algorithm includes a screen blockage identification model, a breakage point reverse reasoning model, and a multi-point breakage automatic arrangement algorithm; wherein, the multi-point breakage automatic arrangement algorithm is used to arrange the breakage operation sequence according to the spatial distribution of the identified multiple breakage points and the principle of shortest path or minimum time.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention integrates a dual sensing device, LiDAR and a 3D camera, and uses AI video analysis technology to fuse multi-source data: LiDAR can quickly acquire high-precision three-dimensional point cloud data, while the 3D camera, combined with AI video analysis, can accurately identify the outline and location of blockage materials under complex working conditions. The two complement each other and effectively overcome the limitations of single sensing devices in harsh environments such as dust and light changes, greatly improving the accuracy and positioning precision of blockage material identification.

[0018] 2. This invention utilizes a combination of LiDAR and a 3D camera to achieve automatic 3D scanning, video acquisition, and AI-powered intelligent recognition of clogged materials. It can quickly acquire the location information of all clogged materials without manual intervention. Simultaneously, through optimal impact sequence planning, it avoids ineffective movement and repeated impacts. Combined with collision-free trajectory planning, the breaker can sequentially complete the crushing of all materials according to the principle of optimal efficiency or distance, significantly shortening the single-time blockage handling time, quickly restoring the grid's flow, effectively preventing production stoppages caused by blockages, and ensuring the continuous operation of the material conveying system. Furthermore, through iterative crushing steps, it can automatically detect the crushing effect and perform secondary crushing, avoiding secondary blockages and further improving operational efficiency.

[0019] 3. This invention uses LiDAR to collect raw point cloud data and combines it with AI analysis of 3D video images from a 3D camera to accurately segment and identify blockage materials and calculate their 3D spatial coordinates. By unifying the coordinate space, the material coordinates are transformed to the robotic arm's base coordinate system, ensuring the accuracy of the crushing target point coordinates. This allows the breaker to accurately strike the core position of the material, avoiding impact deviation, reducing the number of additional strikes, and improving the stability of the crushing effect. At the same time, through closed-loop feedback control of the breaker's chisel, a preset dynamic working gap is maintained, further ensuring crushing accuracy and preventing damage to the grid.

[0020] 4. This invention achieves full automation of the crushing operation, eliminating the need for operators to enter hazardous work areas or focus on manual operation for extended periods, effectively avoiding personal safety risks associated with harsh mining environments and material collapses. Through safety interlocking steps, the work area is monitored in real time. If personnel or mobile equipment intrude into the emergency stop area, an emergency stop command is immediately triggered, interrupting the operation and raising the breaker to a safe position, completely eliminating equipment collisions and personnel injuries. Furthermore, the collision-free trajectory planning effectively avoids collisions between the robotic arm / breaker, the breaker, and the screen and surrounding equipment, reducing the risk of equipment damage and further improving operational safety.

[0021] 5. This invention eliminates the need for manual material identification, position determination, trajectory control, and effect detection. Through the synergistic effect of LiDAR, 3D camera, and AI video analysis, it achieves fully automated control, significantly reducing the labor intensity of operators and avoiding errors caused by fatigue. At the same time, the system automatically completes all core operations through preset algorithms, without relying on the operator's experience. Even novices can quickly put it into use, ensuring the consistency of crushing operation quality in different scenarios.

[0022] 6. This invention features a manual intervention mode. When encountering special working conditions (such as irregular materials or temporary equipment failures), operators can manually control the equipment through a remote or local control panel, improving the system's flexibility and adaptability. At the same time, each step of the system uses mature and reliable algorithms and hardware, resulting in a simple structure that is easy to install and debug. It can be adapted to gantry crushers of different specifications and has a wide range of applications.

[0023] 7. Achieving Closed-Loop Control and Continuous Optimization: This invention forms a closed-loop control for crushing operations through an iterative process of "scanning-collection-identification-crushing-re-scanning," which can automatically detect the crushing effect and promptly replenish the crushed parts. At the same time, through trajectory planning optimization and impact sequence optimization, parameters can be dynamically adjusted according to the actual operation scenario to achieve continuous optimization of crushing operations and further improve operational performance. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a system diagram of the present invention. Detailed Implementation

[0026] The specific embodiments will be described in detail below in conjunction with the working principle of the present invention. Those skilled in the art will understand that the following embodiments are used to explain the present invention, and not to limit its scope of protection.

[0027] The present invention provides a fully automatic control system for a gantry-type crusher, the core of which lies in constructing an automated operation process that integrates environmental perception, intelligent decision-making and precise control.

[0028] Example 1: A complete work cycle of this system in fully automatic mode includes the following steps: S1. Sensing and Localization: After the system starts up, the lidar and 3D camera, fixedly installed on the gantry truss beam, start simultaneously to collaboratively perceive the mine bin grid area below. The lidar emits a laser beam at a certain frequency (e.g., 10Hz) and receives the echo, acquiring massive amounts of 3D point data (raw point cloud data) on the surface of the grid area in real time; the 3D camera simultaneously acquires 3D video image data of the grid area, providing a data source for subsequent AI video analysis. The two complement each other to ensure comprehensive perception in complex environments.

[0029] S2. Target Recognition and Extraction: The raw point cloud data and 3D video image data collected by S1 are fused to segment and identify the clogging material. Specifically, this step can be broken down as follows: S21. Data Preprocessing: First, statistical filtering or radius filtering is applied to the original point cloud to remove discrete outliers caused by dust, splashing water mist, or sensor noise. Median filtering is applied to the 3D video image data acquired by the 3D camera to remove noise. At the same time, a rectangular correction algorithm is used to deal with lens distortion. Through steps such as extracting different color channels, edge detection, contour finding, and polygon ripple removal, a rectangular region that meets the grid sieve characteristics is found and corrected to remove redundant information caused by imaging distortion and environmental interference.

[0030] S22. Multi-source data fusion and plane fitting: Since the grid is usually a roughly planar structure, a random sampling consistency plane fitting algorithm is used to identify and fit a mathematical model representing the grid plane from the denoised point cloud and the corrected 3D video image data respectively. Ax + By + Cz + D = 0. By cross-validating multi-source data, the accuracy of the planar model is optimized, and the initial separation of the background (grid) and the foreground (blocking material) is achieved.

[0031] S23. Material Target Segmentation and Clustering: A preset height threshold (e.g., 0.3m) is set. All point clouds and image pixels whose distances exceed this threshold from the fitted grid plane in S22 are considered as potential blockage material data sets. Then, a density-based spatial clustering algorithm is used to perform cluster analysis on these data. Combined with the grid blockage identification model in AI video real-time analysis, densely connected data points in space are grouped together, thereby identifying each physically independent material pile or large piece of material as an independent data cluster, i.e., a material target.

[0032] S24. Target Coordinate Calculation: For each identified material target data cluster, calculate its centroid coordinates in three-dimensional space and estimate its physical volume. Using a breakpoint reverse reasoning model, noise, outliers, and other useless information are eliminated to filter out effective blockage targets. The system sets a blockage volume threshold, classifying only material targets with volumes greater than this threshold as effective blockage targets. Finally, the centroid coordinates of the effective target (which can be appropriately offset, such as vertically upwards to correspond to the optimal impact point) are output as the three-dimensional spatial coordinates of the material target in the lidar coordinate system or 3D camera coordinate system.

[0033] S3, Coordinate Space Unification: There are fixed spatial geometric relationships between the lidar coordinate system, the 3D camera coordinate system, and the base coordinate system of the crusher's robotic arm. These relationships can be determined through the calibration process during system installation and stored as fixed coordinate transformation matrices (including rotation and translation). The system uses the corresponding transformation matrices to uniformly transform the coordinates of all material targets obtained in S2 in the lidar coordinate system or the 3D camera coordinate system to the robotic arm's base coordinate system, generating the coordinates of the crushing target point that the end of the breaker hammer (chisel rod) needs to reach.

[0034] The transformation of image pixel coordinates from 3D camera images to the world coordinate system (the robotic arm's base coordinate system is an application of the world coordinate system) requires two steps: The first step is to transform the pixel coordinates to the camera coordinate system:

[0035] Multiplying both sides by K inversely leads to the following derivation:

[0036] The second step is to transform from the camera coordinate system to the world coordinate system:

[0037] Multiply the equation by It can be deduced that:

[0038] S4. Motion Planning and Sequence Generation: After receiving the coordinates of all target points, the system begins to plan a safe and efficient movement for the hydraulic breaker.

[0039] S41. Inverse kinematics solution: For each target point, based on the end effector posture required during the operation of the hydraulic breaker, the inverse kinematics model of the robotic arm is used to calculate the angle of rotation required for each joint to drive the robotic arm to reach the target point.

[0040] S42. Trajectory Planning and Collision Detection: To address the potential obstacle obstruction problem during robotic arm movement, a specialized motion control algorithm is employed to achieve linear movement. The core formula is as follows: First, calculate the error propagation Jacobian matrix between the joint space and the end effector:

[0041] The deviation between the expected space and the actual space at the end:

[0042] The final joint angle correction amount is obtained, enabling precise trajectory adjustment:

[0043] The system plans a smooth motion trajectory in joint space or Cartesian space, while performing simulated collision detection to ensure no collisions throughout the process and maintain a safe distance.

[0044] In addition, the system uses a multi-point crushing automatic arrangement algorithm to sort all target points according to the shortest path or minimum time principle based on the spatial distribution of all target points, generating an ordered strike sequence and minimizing the idle travel time of the robotic arm.

[0045] S5. Automatic execution: Once the planning is complete, the system's core controller (such as a PLC or industrial PC) sends commands to the gantry crane traveling mechanism, the servo drives of each joint of the robotic arm, and the hydraulic or electric control system of the hydraulic breaker. Following the trajectory and sequence planned by S4, the robotic arm drives the hydraulic breaker to precisely move to each target point coordinate in sequence. Upon reaching the target point, the controller triggers the hydraulic breaker to execute an automatic breaking action. After completing all strikes in a sequence, the hydraulic breaker can return to a safe standby position.

[0046] Example 2: To further enhance system security and operational thoroughness, this system also integrates the following functions: To further enhance system security and operational thoroughness, this system also integrates the following functions: S0, Safety Interlock: Before and during the automatic operation cycle (S1-S5), an independent safety subsystem operates continuously. This subsystem monitors the gantry's movement area in real time through a sensor network deployed around the work area, including laser scanners, safety light curtains, and cameras. If personnel or unauthorized equipment are detected entering the preset emergency stop zone, the subsystem immediately sends a highest-priority emergency stop command to the crusher's core controller. The controller responds to the command, immediately interrupting all current motion commands and controlling the robotic arm to lift the breaker hammer to an absolutely safe stopping position away from the grid and interference zone until the alarm is cleared.

[0047] Iterative Fragmentation: After completing one round (S5) of automatic strikes, the system does not immediately terminate the mission but automatically returns to step S1, restarting the LiDAR and 3D camera to scan and acquire video of the grid area. Through further processing and planning in S2-S4, it can identify the area after the previous strike: a) Materials that have been completely broken up and no longer constitute a blockage will not be identified as valid targets; b) Materials that have become smaller but are still causing blockages will be re-identified and new target points will be generated; c) Newly exposed or previously unblocked blockage material.

[0048] Based on the new identification results, the system will automatically plan and execute the next round of crushing operations. This process will iterate until the system can no longer identify any valid blockage material target—that is, the screen is cleared—or the system reaches the maximum preset safe number of iterations, at which point it will automatically stop and report that the operation is complete or requires manual inspection.

[0049] Example 3: This system offers a flexible manual intervention mode. When the operator switches the system to manual mode via a remote workstation or local control panel: Operators can send direct motion commands via joystick, button, or touchscreen interface.

[0050] The system interprets these advanced instructions into real-time, continuous, and adjustable control quantities for the speed and direction of the servo motors of each joint of the robotic arm, as well as the hydraulic pressure / current of the hydraulic breaker's telescopic cylinder, thereby achieving precise remote control of the position and movement of the hydraulic breaker.

[0051] In manual mode, the operating interface displays the robotic arm's posture, joint angles, breaker pressure, system alarm information, and real-time video footage captured by a 3D camera in real time, ensuring that the operator can work safely with full knowledge.

[0052] Example 4: During automated crushing operations (S5), the system performs precise closed-loop control of the hammer's impact depth to protect the grid and optimize the impact effect. A high-precision displacement sensor is installed on the intelligent telescopic cylinder of the hydraulic breaker to detect the extension length of the chisel in real time.

[0053] The system calculates a theoretical safe length value in real time based on the known grid plane equation (from S22) and the coordinates of the current target point. This value represents the length at which the end of the drill bit just contacts or slightly penetrates the material, but still maintains a preset dynamic gap with the grid surface.

[0054] During the impact process, the controller continuously compares the extended length of the drill rod with the theoretical safe length. By dynamically adjusting the control signal of the proportional valve or servo motor driving the intelligent telescopic cylinder, the controller ensures that the end of the drill rod maintains the aforementioned dynamic gap with the grid surface during the actual impact. This ensures that the crushing force is effectively transmitted to the material while preventing the drill rod from directly impacting the hard grid, thus protecting the grid structure and extending the life of the drill rod.

[0055] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "connected" and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0056] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A fully automatic control system for a gantry-type crusher, characterized in that, Includes the following steps: S1. Perception and Positioning: A lidar and a 3D camera are fixedly installed on the crossbeam of the gantry truss to perform 3D scanning and video acquisition of the grid area of ​​the mine below, and to obtain the original point cloud data and 3D video image data of the grid area in real time. Among them, the 3D video image data acquired by the 3D camera is used for AI video analysis, which complements the original point cloud data acquired by the lidar to improve the accuracy of perception. S2. Target Recognition and Extraction: The raw point cloud data and the three-dimensional video image data collected by the 3D camera are fused and processed. The AI ​​video analysis algorithm is used to segment and identify the material targets to be crushed that are blocked above the grid, and to calculate the three-dimensional spatial coordinates of each material target in the lidar coordinate system or the 3D camera coordinate system. S3. Coordinate Space Unification: Based on the preset fixed coordinate transformation relationship between the gantry truss and the lidar and 3D camera, the three-dimensional spatial coordinates of each material target are uniformly transformed to the base coordinate system of the crusher robotic arm, generating the corresponding crushing and impact target point coordinates. S4. Motion Planning and Sequence Generation: Based on the coordinates of the target point of the breaker, a collision-free motion trajectory is planned for the breaker, and an ordered strike sequence is generated by combining efficiency or distance principles. S5. Automatic execution: Controls the gantry truss crusher to drive the robotic arm and breaker hammer, and moves them to each target point in sequence according to the above motion trajectory and impact sequence, and performs automatic crushing actions.

2. The fully automatic control system for a gantry-type crusher according to claim 1, characterized in that, Step S2 specifically includes: S21. Filter and denoise the original point cloud data to remove discrete noise points; perform median filtering and rectangle correction on the 3D video image data acquired by the 3D camera to eliminate redundant information caused by imaging distortion and environmental interference, and ensure the accuracy of image data. S22. The random sampling consistency plane fitting algorithm is adopted to fit the mathematical model of the grid plane from the denoised point cloud and the corrected 3D video image data respectively. The accuracy of the plane model is improved by cross-validation of multi-source data. S23. Point cloud and image pixels that are more than a preset height threshold away from the mathematical model of the grid plane are taken as the data set to be processed. The set is segmented by a density-based spatial clustering algorithm. Combined with the grid blockage identification model in AI video real-time analysis, each connected data cluster is identified as an independent material target. S24. Calculate the physical volume and three-dimensional centroid of the data cluster corresponding to each material target. Filter effective blockage targets through the reverse reasoning model of the break point. Determine the centroid coordinates of the material targets whose volume is greater than the preset blockage volume threshold as the three-dimensional spatial coordinates of the material target in the corresponding coordinate system.

3. The fully automatic control system for a gantry-type crusher according to claim 1, characterized in that, In step S4, planning a collision-free motion trajectory for the hydraulic breaker specifically includes: S41. For the coordinates of each breaking and impact target point, and in combination with the attitude requirements of the end effector of the hydraulic breaker, the target angles of each joint are solved using the inverse kinematics model of the robotic arm. S42. Perform trajectory interpolation in joint space or Cartesian space. During the planning process, use simulation collision detection to ensure that the breaker rod and the robotic arm body maintain a safe distance from the grid and the surrounding environment.

4. The fully automatic control system for a gantry-type crusher according to claim 1, characterized in that, Before step S1, the following is also included: S0. Safety Interlock: The active area of ​​the gantry crusher is monitored in real time by a sensor network deployed in the work area; when personnel or mobile equipment are detected to have entered the preset emergency stop area, an emergency stop command is immediately sent to the core controller of the crusher; the core controller responds to the emergency stop command, interrupts any automatic crushing operation that is being performed or about to be performed, and controls the breaker hammer to be raised to a safe stopping position.

5. The fully automatic control system for a gantry-type crusher according to claim 1, characterized in that, After completing one round of attacks in step S5, an iterative breaking step is also included: Return to step S1 and scan the grid area again, acquire video, and identify the target using LiDAR and 3D camera; If a valid blockage target is identified again, steps S2 to S5 are repeated for the next round of crushing until the grid area is clear or the maximum number of iterations set by the system is reached.

6. The fully automatic control system for a gantry-type crusher according to claim 1, characterized in that, It also includes a manual intervention mode, which, when switched to, allows for: It receives direct motion commands from a remote operating station or local operating panel, and parses the motion commands into real-time control quantities for the servo motors of each joint of the robotic arm and the telescopic cylinder of the breaker, so as to realize continuous and adjustable speed manual control of the position and movement of the breaker. Meanwhile, in manual intervention mode, all key status parameters of the crusher, alarm information, and real-time video footage from the 3D camera are fed back to the operation interface.

7. The fully automatic control system for a gantry-type crusher according to claim 1, characterized in that, In step S5, when controlling the breaker hammer to perform automatic crushing operations: The extension length of the chisel is detected in real time by a high-precision displacement sensor installed on the intelligent telescopic cylinder of the hydraulic breaker. The detected actual length value is compared with the theoretical safe length value calculated based on the grid plane and target point coordinates to form a closed-loop feedback; By dynamically adjusting the control signal of the intelligent telescopic cylinder, a preset dynamic working gap is maintained between the end of the chisel and the grid surface during the actual crushing process.

8. The fully automatic control system for a gantry-type crusher according to claim 2, characterized in that, The specific process for rectangular correction is as follows: Median filtering is used to denoise the 3D video image data acquired by the 3D camera; Extract different color channels (BGR) and detect rectangles separately; Canny edge detection or multi-threshold binarization is applied to each color channel; Use the findContours function to find image contours; The approxPolyDP function is used to remove minor undulations in the polygon outline; Select quadrilaterals with larger areas and convex shapes; Determine if the cosine of the angle between any two adjacent lines of the quadrilateral is less than 0.

3. If so, the quadrilateral is determined to be the target rectangle, and the correction is completed.

9. The fully automatic control system for a gantry-type crusher according to claim 1, characterized in that, The AI ​​video analysis algorithm includes a screen blockage identification model, a breakage point reverse reasoning model, and a multi-point breakage automatic scheduling algorithm. The multi-point breakage automatic scheduling algorithm is used to schedule the breakage operation sequence according to the shortest path or minimum time principle based on the spatial distribution of multiple identified breakage points.

Images