A columnar drill rod box mechanical hand inspection control method

Abstract

The application discloses a columned drill pipe box mechanical arm inspection control method, which comprises the following steps: a central controller receives a knob combination control signal and judges a current task mode; when a trigger condition is met, the mechanical arm is controlled to globally scan a drill pipe box, the distribution state of drill pipes in each bin is acquired, and a target bin is selected based on the distribution state and the task mode; the moving direction is judged and the moving speed curve is planned according to the current position of the mechanical arm and the theoretical position of the target bin; the local perception unit arranged on the mechanical arm body is used to locally detect the target bin, the accurate position information and the quantity state information of the drill pipes in the target bin are acquired, the mechanical arm is controlled to adjust the end pose and execute the taking and placing operation; after a single taking and placing operation is completed, the drill pipe quantity of the corresponding bin is synchronously updated, and the mechanical arm decides the target bin of the next operation according to the updated task target. The application improves the efficiency, accuracy and reliability of the drill pipe storage and taking operation.

Description

Technical Field

[0001] This invention relates to the field of drill rod conveying control technology for coal mine drilling rigs, specifically to a method for inspection and control of a split-type drill rod box robotic arm. Background Technology

[0002] In mining and other engineering fields, drill pipe is a core drilling consumable, and its storage, transportation, and management efficiency directly affects the overall construction progress. Split-type drill pipe boxes are widely used in centralized drill pipe storage scenarios due to their large storage capacity, high space utilization, and ease of classification and management. However, current drill pipe inspection and retrieval operations using split-type drill pipe boxes still face many technical bottlenecks, and traditional operating methods are insufficient to meet the demands of modern engineering for efficient, precise, and safe operations.

[0003] In existing technologies, the inspection and retrieval of drill pipe boxes mostly rely on manual operation or semi-automatic control methods. In the manual operation mode, operators need to closely observe the distribution, quantity, and position of drill pipes in each compartment of the drill pipe box, and then operate the robotic arm to complete the retrieval and placement. This is not only labor-intensive and inefficient, but also prone to problems such as missed inspections and incorrect inspections due to human visual fatigue and judgment errors, or drill pipe collisions and damage to the robotic arm due to misalignment during retrieval and placement.

[0004] While semi-automatic control methods have partially replaced manual observation, they still have significant drawbacks: First, the global scanning accuracy is insufficient, as they often use single-point scanning or simple image acquisition methods, making it difficult to fully obtain the true distribution status of drill rods in each compartment, which can easily lead to misselection of target compartments; Second, the local positioning and quantity detection methods are limited, lacking a precise perception mechanism that integrates multiple sensors, making it impossible to accurately obtain the precise location, layer number, and quantity information of drill rods, thus affecting the success rate of retrieval and placement operations. Summary of the Invention

[0005] In order to solve the above-mentioned technical problems, this application proposes the following technical solution:

[0006] This application provides a method for inspecting and controlling a split-type drill pipe box robotic arm, including:

[0007] After receiving the control signals from the remote control, the central controller judges the current task mode according to the preset correspondence. When the preset inspection trigger conditions are met, the robot arm starts the inspection operation.

[0008] The robotic arm is controlled to perform a global scan of the drill pipe box to obtain the distribution status of drill pipes in each compartment, and the target compartment is selected based on the distribution status of the drill pipes and the current task mode.

[0009] Based on the current position of the robotic arm and the theoretical position of the target compartment, determine the direction of movement and plan the movement speed curve;

[0010] After the robotic arm moves to the predetermined position of the target compartment according to the speed curve, it uses the local sensing unit set in the robotic arm body to perform local detection on the target compartment and obtain the precise position information and quantity status information of the drill rod in the target compartment.

[0011] Based on the obtained precise position and quantity information of the drill rod, the robot arm is controlled to adjust its end effector posture and perform pick-up and drop-off operations.

[0012] After a single retrieval and placement operation is completed, the drill pipe quantity information of the corresponding compartment is updated synchronously. Based on the updated global drill pipe distribution data, transfer position status and task objectives, the robotic arm decides the target compartment for the next operation.

[0013] In one possible implementation, the robotic arm is controlled to perform a global scan of the drill pipe box to obtain the distribution status of drill pipes in each compartment, and based on the distribution status of the drill pipes and the current task mode, a target compartment is selected, including:

[0014] The robot arm is controlled to move to several predefined global observation points, and a fixed 2D industrial camera is used to perform a global scan of the drill pipe box;

[0015] The distribution status of drill pipes in each compartment is obtained using the YOLO target detection algorithm;

[0016] Based on the distribution of the drill pipe and the current task mode, the target compartment is selected.

[0017] In one possible implementation, the step of using the YOLO target detection algorithm to obtain the distribution state of the drill pipe in each compartment includes:

[0018] The acquired panoramic image is divided into The grid;

[0019] After predicting the probabilities of B bounding boxes and their classes using each grid cell, the results are integrated and output as a single value. tensor;

[0020] By using non-maximum suppression processing, overlapping low-confidence prediction boxes are removed, and the final output is the location information of all grid cells that have been identified as drill pipes.

[0021] In one possible implementation, determining the direction of movement and planning the speed curve based on the current position of the robotic arm and the theoretical position of the target compartment includes:

[0022] Compare the current position of the robotic arm with the theoretical position of the target compartment;

[0023] When the current position of the robotic arm is less than the theoretical position of the target compartment, it moves in the positive direction; when the current position of the robotic arm is greater than the theoretical position of the target compartment, it moves in the negative direction.

[0024] Then, the distance difference between the current position of the robotic arm and the theoretical position of the target compartment is calculated;

[0025] The distance traveled is determined based on the distance difference.

[0026] When the distance difference is greater than 280mm, it is a long-distance movement; when the distance difference is greater than 7mm and less than or equal to 280mm, it is a medium-distance movement; when the distance difference is less than or equal to 7mm, it is a short-distance movement.

[0027] When moving in the positive direction, if it is determined to be a long-distance movement, the acceleration interval is as follows: The deceleration range is When the movement is determined to be of medium distance, the acceleration and deceleration distances are each half of the travel distance, and the travel speed is lower than the high-speed travel speed; when the movement is determined to be of short distance, the movement is carried out at a fixed low speed.

[0028] When moving in the negative direction, if it is determined to be a long-distance movement, the acceleration interval is as follows: The deceleration range is When the movement is determined to be of medium distance, the acceleration and deceleration distances are each half of the travel distance, and the operating speed is lower than the high-speed operating speed; when the movement is determined to be of short distance, the movement is carried out at a fixed low speed, wherein... This is the initial position of the robotic arm. The theoretical position of the target cell.

[0029] In one possible implementation, after the robotic arm moves to a predetermined position in the target compartment according to the speed curve, it uses a local sensing unit installed on the robotic arm body to perform local detection of the target compartment, obtaining precise position and quantity information of the drill rods within the target compartment, including:

[0030] The robotic arm moves to the theoretical position above the target compartment according to the speed curve;

[0031] Local images of the target compartment are captured by an industrial camera mounted on a robotic arm;

[0032] A vision algorithm based on geometric matching is used to identify the actual center coordinates of the drill pipe in the image and compare them with the theoretical center coordinates of the grid to calculate the positional deviation.

[0033] The robot arm is controlled to perform offset compensation based on the position deviation, so that the center of the robot arm end effector is aligned with the center of the drill pipe;

[0034] Then, the laser rangefinder is activated to measure the surface profile of the drill rods in the compartment. Based on the height change steps, the number of drill rod layers and quantities in the current compartment are calculated and updated to the drill rod quantity status information.

[0035] In one possible implementation, a geometric matching-based visual algorithm is used to identify the actual center coordinates of the drill pipe in the image and compare them with the theoretical center coordinates of the storage grid to calculate the positional deviation, including:

[0036] After calibrating the camera and hand-eye coordination, standard drill pipe images were acquired.

[0037] Select a region of interest containing stable features in the standard drill pipe image, and create a geometric matching template based on the region of interest;

[0038] The central controller triggers the camera to acquire images of the target cell.

[0039] After preprocessing the acquired target grid image, template matching is performed to calculate the similarity score between the template and the image edge, and the best matching position is obtained.

[0040] The algorithm outputs matching position transformation parameters, which include the image coordinates and rotation angle of the drill pipe center;

[0041] The central controller calculates the position deviation based on the transformation relationship between the coordinates and the robot's coordinate system.

[0042] In one possible implementation, the formulas for calculating the number of drill pipe layers and the quantity of drill pipes in the current cell based on the height change steps are as follows:

[0043]

[0044]

[0045]

[0046] in, This refers to the number of drill pipe layers. The equivalent optical height difference for a single-layer drill pipe. This represents the average height of the entire effective contour region. This is the original distance value when the warehouse is empty. This is the distance value below one standard drill pipe layer. Let X be the coverage width of a certain platform in the X direction. The diameter of the drill pipe is [diameter]. The gap between adjacent drill pipes. This refers to the number of drill pipes in a single layer.

[0047] In one possible implementation, based on the acquired precise position and quantity information of the drill rod, the robot arm is controlled to adjust its end effector pose and perform pick-and-place operations, including:

[0048] Based on the confirmed number of drill pipes, the robotic arm plans the vertical descent speed curve and descends into the target compartment;

[0049] During the descent, the proximity switch signal installed at the end of the robotic arm and the vertical displacement sensor signal are monitored in real time.

[0050] If a valid signal from the proximity switch is received within the expected height range, the drill pipe is determined to be present, and the gripper is controlled to perform a gripping action.

[0051] If no valid proximity switch signal is received after descending to the preset limit position, the compartment is determined to be empty, the robotic arm is raised and switched to the next target compartment according to the decision logic;

[0052] If abnormal pressure or no signal from the proximity switch is detected during descent or gripping, the robot arm's operation will be stopped immediately and an alarm will be triggered.

[0053] Compared with the prior art, the beneficial effects of this application are as follows:

[0054] This application transforms drill pipe management from a crude, experience-dependent operation into a quantifiable, predictable, and highly precise intelligent process. It eliminates human error and completely avoids problems such as incorrect handling, misplacement, and recording errors caused by operator fatigue, lack of experience, or inattention. Every operation is strictly executed according to the procedure, automatically updating each data column, ensuring stable and reliable quality. Simultaneously, it fundamentally eliminates personal injury, completely freeing operators from the heavy and dangerous drill pipe handling area and preventing serious safety accidents caused by drill pipe rolling, slipping, or crushing.

[0055] This application utilizes a central controller to analyze remote control knob combinations to intelligently determine the task mode. After triggering the inspection operation, the controller moves the robotic arm to multiple predefined global observation points, combining a fixed 2D industrial camera with the YOLO target detection algorithm to complete a global scan of the drill pipe box. This design overcomes the limitations of traditional single-observation-point scanning. By integrating multi-observation-point full-coverage acquisition with a deep learning target detection algorithm, it can accurately identify the distribution of drill pipes in each compartment, ensuring more scientific and accurate target compartment selection based on drill pipe distribution and task mode. This effectively avoids missed or incorrect selections, laying a precise foundation for subsequent operations.

[0056] This application employs a multi-sensor fusion local perception scheme, combining visual positioning and laser ranging, after moving the robotic arm to the predetermined position of the target compartment. Through a geometric matching-based visual algorithm, the actual center coordinates of the drill rod can be accurately identified and positional deviations calculated, achieving precise offset compensation for the robotic arm's end effector. Combined with laser ranging sensors detecting the drill rod's surface contour, the number and quantity of drill rod layers are calculated using height variation steps, ensuring comprehensive and accurate control over the drill rod's condition. This design effectively solves the problems of traditional local perception methods being singular and lacking accuracy, significantly reducing the risks of retrieval / placement failures and drill rod collisions caused by positioning deviations and quantity misjudgments, thus improving the success rate and safety of retrieval / placement operations.

[0057] This application features comprehensive inspection and alarm functions. When abnormalities such as drill rod misalignment or gripping failure are detected, it can immediately stop and trigger an alarm, rather than forcibly continuing operation and causing more serious equipment damage or accidents. The proximity switch at the robotic arm's end effector also provides better feedback of actual signals, preventing collisions. Attached Figure Description

[0058] Figure 1 A flowchart illustrating a method for inspecting and controlling a split-type drill pipe box robot provided in an embodiment of this application;

[0059] Figure 2 A schematic diagram of the geometric relationship of the laser triangulation method provided in the embodiments of this application;

[0060] Figure 3 This is a schematic diagram of the mechanical structure of the split-type drill pipe box robot inspection control provided in the embodiments of this application.

[0061] in, Figure 3 The symbols in the middle are: 1-transfer mechanism, 2-hydraulic proportional control manipulator, 3-split drill pipe box. Detailed Implementation

[0062] The present solution will now be described in conjunction with the accompanying drawings and specific embodiments.

[0063] Figure 1 A flowchart illustrating a method for inspecting and controlling a split-type drill pipe box robot provided in this application embodiment is shown below. Figure 1 The method for inspection and control of a split-type drill pipe box robotic arm in this embodiment includes:

[0064] S101: After receiving the control signal from the combination of toggle switches from the remote controller, the central controller judges the current task mode according to the preset correspondence. When the preset inspection trigger conditions are met, the robot arm starts the inspection operation.

[0065] In this embodiment, since the on-site tasks are controlled by a combination of remote control knobs, the tasks are arranged into an array within the central controller. After receiving the knob control signals from the on-site remote controller, the central controller determines the current task mode based on the knob combination. When a task begins or a drill rod is available at the transfer position, the robotic arm is activated for inspection. The task modes include drill rod loading and drill rod unloading. In drill rod loading mode, the robotic arm retrieves drill rods from the drill rod box compartment and transports them to the transfer position. In drill rod unloading mode, the robotic arm retrieves drill rods from the transfer position and returns them to an unfilled compartment in the drill rod box. Before returning the drill rod, the actual available capacity of the target return compartment must be confirmed through a global coarse inspection and laser ranging verification.

[0066] S102, control the robotic arm to perform a global scan of the drill pipe box, obtain the distribution status of drill pipes in each compartment, and select the target compartment based on the distribution status of drill pipes and the current task mode.

[0067] In this embodiment, each column of drill rods is designed as an array, including the target movement position of each column and the number of drill rods in each column. The robot arm is controlled to move to several predefined global observation points, and a fixed 2D industrial camera performs a global scan of the drill rod box. The YOLO target detection algorithm is used to obtain the distribution status of the drill rods in each compartment.

[0068] YOLO, an object detection algorithm, is a deep learning algorithm capable of simultaneously identifying multiple objects in an image and assigning their locations and categories. Its key features are speed and end-to-end processing. Object detection is a multi-step pipeline: first, a large number of candidate regions are generated in the image; then, each candidate region is classified using a convolutional neural network; finally, the results are post-processed for optimization; and finally, the network directly predicts the bounding boxes and class probabilities of all objects in the image at the output layer. Specifically, it includes: dividing the input image into a... The grid is divided into two parts, and each grid cell makes a prediction: each grid cell predicts B bounding boxes and one class probability. Each bounding box contains 5 predicted values: The coordinates of the center point of the frame relative to the boundary of the grid cell. The width and height of the bounding box are proportional to the entire image. The confidence score indicates how likely the box is to contain an object, and the accuracy of the predicted box relative to the ground truth box. Finally, the network integrates and outputs a... The tensor of the array is used. Since multiple grid cells may predict the same object, many overlapping predicted boxes will eventually be generated. Non-maximum suppression is then used to remove duplicates. First, all boxes with confidence scores below a threshold are discarded. Among the remaining boxes, the box with the highest confidence score is selected as the survivor. The overlap area between this survivor box and all other boxes is calculated, and duplicate predicted boxes are removed. This process is repeated for the remaining boxes until all boxes have been processed.

[0069] S103 determines the direction of movement and plans the movement speed curve based on the current position of the robotic arm and the theoretical position of the target compartment.

[0070] In this embodiment, the current position of the robotic arm is compared with the theoretical position of the target compartment to determine whether the robotic arm is moving in a positive or negative direction. When the current position of the robotic arm is less than the theoretical position of the target compartment, it is considered positive movement; when the current position is greater than the theoretical position, it is considered negative movement. Then, the distance difference between the current position of the robotic arm and the theoretical position of the target compartment is calculated. Based on this distance difference, the movement distance is determined: when the distance difference is greater than 280mm, it is considered long-distance movement; when the distance difference is greater than 7mm and less than or equal to 280mm, it is considered medium-distance movement; and when the distance difference is less than or equal to 7mm, it is considered short-distance movement. In positive direction movement, when it is determined to be a long-distance movement, the acceleration interval is initiated. The deceleration range is When the movement is determined to be of medium distance, the acceleration and deceleration distances are each half of the travel distance, and the operating speed is lower than the high-speed operating speed. When the movement is determined to be of short distance, it operates at a fixed low speed. In negative directional movement, when the movement is determined to be of long distance, the acceleration interval is... The deceleration range is When the movement is determined to be of medium distance, the acceleration and deceleration distances are each half of the travel distance, and the operating speed is lower than the high-speed operating speed; when the movement is determined to be of short distance, the movement is carried out at a fixed low speed, wherein... This is the initial position of the robotic arm. The theoretical position of the target cell.

[0071] S104, after the robotic arm moves to the predetermined position of the target compartment according to the speed curve, it uses the local sensing unit set in the robotic arm body to perform local detection on the target compartment and obtain the precise position information and quantity status information of the drill rod in the target compartment.

[0072] In this embodiment, the robotic arm moves to a theoretically higher position than the target compartment according to the speed curve. A local image of the target compartment is captured by an industrial camera mounted on the robotic arm. Since the robotic arm is driven by a proportional valve group powered by a variable displacement motor, there is a start-up and stop delay. To ensure accurate positioning, after reaching the pre-stop position, it is determined whether the robotic arm's position is within the upper or lower range of the target position. The positioning is then repeated in both directions. The industrial camera is used to capture the drill pipe position, and a geometric matching algorithm is used for offset compensation to ensure the robotic arm stops at the target position, guaranteeing precise alignment between the gripper center and the drill pipe compartment center.

[0073] In this embodiment, the method for offset compensation judgment using a geometric matching algorithm from a visual database specifically includes:

[0074] Phase 1: Offline Training Setup. First, camera calibration is performed to eliminate lens distortion and obtain accurate intrinsic parameters. Second, hand-eye calibration is performed to determine the transformation relationship between the camera coordinate system and the robot arm base coordinate system (eye outside the hand and eye on the hand). Then, using a sample drill rod with a known good position and angle, a clear image is captured under suitable lighting conditions. Regions with drill rod features are selected within the image, choosing stable, unique, and high-contrast features to reduce computation and improve accuracy and speed. Based on the selected ROI, a geometric matching algorithm is used to generate a template for searching, extracting the contour edges and gradient directions of the object. A matching score threshold is set, indicating a confidence threshold; results exceeding this threshold are considered valid. The number of pyramid levels is used to accelerate the search. The scale range specifies the scaling range for the drill rod size.

[0075] Phase Two: Online Operation. First, the central controller sends a trigger signal to control the industrial camera to capture an image of the current field of view. The acquired image is then processed with filtering, binarization, and edge enhancement to improve image quality. When a new image is input, the algorithm similarly extracts the image edges, calculates the similarity score between the template edge and the image edges, and uses a correlation score to evaluate the matching quality. When a position is found where the template edge and the image edge have the highest overlap and are aligned in the same direction, the drill rod is considered found. Once the best match is found, the algorithm returns one or more transformation parameters. These parameters describe how to move the template into the current image to coincide with the drill rod. This transformation parameter is the positional information, including (Y): the row coordinates (pixels) of the drill rod center point in the image coordinate system; (X): the column coordinates (pixels) of the drill rod center point in the image coordinate system. The coordinates are: the rotation angle of the drill rod relative to the template; and the scaling ratio of the drill rod. The obtained coordinates are sent to the central controller via Ethernet, and then compared with the current position of the robot arm within the central controller for offset compensation.

[0076] Then, the laser rangefinder is activated to measure the surface profile of the drill rods in the compartment. Based on the height change steps, the number of drill rod layers and quantities in the current compartment are calculated and updated to the drill rod quantity status information.

[0077] See Figure 2 In this embodiment, a laser triangulation rangefinder is fixedly installed above the cell, and its laser line is projected vertically downwards onto the drill rod surface. When the cell is empty, the laser line hits the bottom of the cell, and the maximum distance measured is [value missing]. When a layer of drill rods is inserted, the laser beam hits the surface of the topmost drill rod, and the measured distance value will be reduced by the diameter of one drill rod. With each additional layer of drill pipe, the sensor reading decreases by approximately By analyzing the distance-position profile curve through single-point measurements, step-like transitions were identified, with each step representing a layer of drill pipe.

[0078] The measurement principle involves measuring the height profile of the drill pipe surface in the storage compartment, and then using the step changes in the profile to infer the number of drill pipe layers. Specifically, this includes: fixing a linear laser triangulation sensor to a robotic arm, ensuring the laser line completely covers the width of the storage compartment, and recording the original distance value measured by the sensor in an empty compartment state. This is the reference plane for measurement. Single-layer calibration: Place a standard drill pipe layer and record the measured values. The distance change caused by a single layer of drill pipe was calculated. This is approximately equal to the diameter of the drill pipe. Establish the height-layer correspondence:

[0079]

[0080] in, For the number of floors, The current measured height, It is the equivalent optical height difference of a single-layer drill pipe.

[0081] The sensor is then triggered to acquire a complete laser profile, which is an array: [position index i, distance value...]. The raw data needs processing to clearly represent the steps. Median filtering is used to remove signal glitches and noise. Walls or shadows may generate invalid data, which is discarded based on a threshold. Transform the height value: Height value Directly represents the height of the drill pipe surface relative to the bottom of the chamber, empty space The value is 0, at the location of the drill pipe. The number of layers is calculated using the mean method, and the average height of the entire effective contour area is calculated. The formula for calculating the number of layers is as follows:

[0082]

[0083]

[0084]

[0085] in, This refers to the number of drill pipe layers. The equivalent optical height difference for a single-layer drill pipe. This represents the average height of the entire effective contour region. This is the original distance value when the warehouse is empty. This is the distance value below one standard drill pipe layer. Let X be the coverage width of a certain platform in the X direction. The diameter of the drill pipe is [diameter]. The gap between adjacent drill pipes. This represents the number of drill pipes per layer. The calculated number of layers N, along with the possible number of pipes per layer, is sent to the central controller via the communication interface.

[0086] In addition, the basic formula for triangulation (which has been processed internally by the sensor to provide a direct distance value) is as follows:

[0087]

[0088] when When the angle is close to 90°, meaning the optical axis of the receiving lens is roughly perpendicular to the object's surface, and the displacement x is small:

[0089]

[0090] Where D is the measured height change, i.e., the displacement of the object's surface along the laser direction; x is the pixel offset of the laser spot on the camera's imaging chip (measured from the center of the optical axis); and L is the baseline length, the horizontal distance between the laser spot at the laser and the center of the camera lens. The laser incident angle is the angle between the laser beam and the normal to the object's surface. The effective focal length of the camera lens; The camera receiving angle is the angle between the camera's optical axis and the normal to the object's surface.

[0091] S105, based on the obtained precise position and quantity information of the drill rod, controls the robot arm to adjust the end effector posture and perform pick-up and drop operations.

[0092] In this embodiment, based on the determined number of drill rods in the compartment and the vertical displacement sensor of the robot arm, the robot arm can determine the descent speed range based on the quantity value, and initiate acceleration, high-speed deceleration, and deceleration to prevent the robot arm from colliding with the drill rods due to excessive speed. The descent stops when the robot arm reaches the height range of the drill rod position and the proximity switch at the end of the robot arm provides a signal feedback. If the proximity switch does not provide a signal feedback when the target drill rod height is reached, the robot arm will continue to descend until the proximity switch provides a signal or the proximity switch stops when the limit is exceeded. Here, the proximity switch feedback signal serves three purposes: first, to provide feedback that the position has reached and the drill rod has been detected; second, to assist the robot arm in stopping at the correct position; and third, to cooperate with the vertical displacement sensor to determine that the compartment is indeed empty. After the actual position is reached, the actual number of drill rods in the compartment is recalculated and compared with the number determined by the laser ranging in step four for calibration.

[0093] After the proximity switch indicates that there is a drill rod, the robotic arm gripper clamps the drill rod. Once the clamping pressure reaches the set value, it is determined that the drill rod is clamped and can be lifted to remove it from the drill rod box. After the drill rod is raised to the correct position, the number of drill rods in this column is reduced by one.

[0094] If the proximity switch does not provide feedback when the robot arm descends beyond the limit, it is determined that the compartment is empty. The robot arm rises and switches to the next column to descend and grab the object. If the compartment is empty, the robot arm continues to switch columns until it is determined that there is a drill rod in the compartment or all compartments are empty after switching and judging.

[0095] If the expected force feedback is not detected within a certain distance during the descent, such as missing drill rod, incorrect height, abnormally increased pressure, drill rod tilting and jamming, or no signal from the proximity switch and abnormally increased pressure, the descent should be stopped immediately and an alarm should be triggered.

[0096] S106 After a single pick-up and drop-off operation is completed, the drill pipe quantity information of the corresponding compartment is updated synchronously. Based on the updated global drill pipe distribution data, transfer position status and task objectives, the robot arm decides the target compartment for the next operation.

[0097] In this embodiment, after the robotic arm picks up a drill rod from the drill rod compartment, it determines whether to transport the drill rod based on whether the transfer position is empty. If there is a drill rod in the transfer position, the transport is paused until the transfer position becomes empty and the drill rod is in place before resuming transport. When the transfer position becomes empty, the robotic arm picks up the drill rod and continues transporting. The robotic arm moves according to the decided running direction and speed plan, placing the drill rod in the transfer position. At the transfer position, the robotic arm's vertical displacement sensor determines the descent position, and the proximity switch for drill rod detection at the transfer position determines whether the drill rod is in place. The robotic arm's movement direction is determined based on the number of drill rods remaining in the compartment from the previous drill rod retrieval. If there are remaining drill rods in the previous compartment, the robotic arm moves to that compartment again. If the compartment is empty, the robotic arm moves to the next compartment according to the decision of the central controller, and then repeats the above process. When the robotic arm picks up or puts back a drill rod from the drill rod column, the data of each drill rod array is updated and saved.

[0098] The movement method for the lower drill pipe mode is the same as that for the upper drill pipe mode. The difference is that the lower drill pipe mode involves returning the drill pipe to the drill pipe compartment, which is the reverse process of the upper drill pipe mode. The robot arm first needs to perform a global inspection of the drill pipe column to determine the number of drill pipes in each compartment. Empty compartments are skipped, and the robot arm starts from the compartments that are not full. The robot arm descends to actually detect the number of drill pipes in the unfilled compartments, corrects the number of drill pipes calculated by the laser measurement, and waits for the lower drill pipes to be in place at the transfer position. After the lower drill pipes are in place, the central controller issues a movement command, and the robot arm moves to the transfer position to grab the drill pipes and return the drill pipes in the unfilled compartments. After the compartment is full, the robot arm will switch to the next column to actually detect, ensuring that the laser detection is accurate and that the return of drill pipes is safe and will not cause the robot arm to bump into anything. After all columns are full, the robot arm stops the next action and issues an alarm to remind the worker.

[0099] See Figure 3This embodiment of the split-type drill pipe box robotic inspection control system includes a split drill pipe box 3, a hydraulic proportional control robotic arm 2, a sensing system, and a central controller. The split drill pipe box is composed of multiple independent compartments arranged in a matrix. The hydraulic proportional control robotic arm, mounted on the walking mechanism, can move above and to the side of the drill pipe box; its power is provided by a variable displacement hydraulic motor, and a proportional valve group controls the flow and pressure. The sensing system includes a fixed industrial camera for overall coarse inspection, an industrial camera mounted on the robotic arm for local fine positioning, a laser rangefinder for quantity detection, a proximity switch for position detection, a displacement sensor for measuring displacement, and a pressure sensor for detecting clamping force. The central controller integrates a motion control card, a PLC, and a host industrial computer, responsible for data processing, process execution, task scheduling, and motion command issuance.

[0100] In this embodiment, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, the simultaneous existence of A and B, or the existence of B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0101] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0102] The above description is merely a specific embodiment of this application. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the protection scope of this application. The protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A method for inspection and control of a split-type drill pipe box robotic arm, characterized in that, include: After receiving the control signals from the remote control, the central controller judges the current task mode according to the preset correspondence. When the preset inspection trigger conditions are met, the robot arm starts the inspection operation. The robotic arm is controlled to perform a global scan of the drill pipe box to obtain the distribution status of drill pipes in each compartment, and the target compartment is selected based on the distribution status of the drill pipes and the current task mode. Based on the current position of the robotic arm and the theoretical position of the target compartment, determine the direction of movement and plan the movement speed curve, including: Compare the current position of the robotic arm with the theoretical position of the target compartment; When the current position of the robotic arm is less than the theoretical position of the target compartment, it moves in the positive direction; when the current position of the robotic arm is greater than the theoretical position of the target compartment, it moves in the negative direction. Then, the distance difference between the current position of the robotic arm and the theoretical position of the target compartment is calculated; The distance traveled is determined based on the distance difference. When the distance difference is greater than 280mm, it is a long-distance movement; when the distance difference is greater than 7mm and less than or equal to 280mm, it is a medium-distance movement; when the distance difference is less than or equal to 7mm, it is a short-distance movement. When moving in the positive direction, if it is determined to be a long-distance movement, the acceleration interval is as follows: The deceleration range is When the movement is determined to be of medium distance, the acceleration and deceleration distances are each half of the travel distance, and the travel speed is lower than the high-speed travel speed; when the movement is determined to be of short distance, the movement is carried out at a fixed low speed. When moving in the negative direction, if it is determined to be a long-distance movement, the acceleration interval is as follows: The deceleration range is When the movement is determined to be of medium distance, the acceleration and deceleration distances are each half of the travel distance, and the operating speed is lower than the high-speed operating speed; when the movement is determined to be of short distance, the movement is carried out at a fixed low speed, wherein... This is the initial position of the robotic arm. The theoretical position of the target storage cell; After the robotic arm moves to the predetermined position of the target compartment according to the speed curve, it uses the local sensing unit set in the robotic arm body to perform local detection on the target compartment and obtain the precise position information and quantity status information of the drill rod in the target compartment. Based on the obtained precise position and quantity information of the drill rod, the robot arm is controlled to adjust its end effector posture and perform pick-up and drop-off operations. After a single retrieval and placement operation is completed, the drill pipe quantity information of the corresponding compartment is updated synchronously. Based on the updated global drill pipe distribution data, transfer position status and task objectives, the robotic arm decides the target compartment for the next operation.

2. The method for inspection and control of a split-type drill pipe box robotic arm according to claim 1, characterized in that, The robotic arm is controlled to perform a global scan of the drill pipe box, acquire the distribution status of drill pipes in each compartment, and select a target compartment based on the distribution status of the drill pipes and the current task mode, including: The robot arm is controlled to move to several predefined global observation points, and a fixed 2D industrial camera is used to perform a global scan of the drill pipe box; The distribution status of drill pipes in each compartment is obtained using the YOLO target detection algorithm; Based on the distribution of the drill pipe and the current task mode, the target compartment is selected.

3. The method for inspection and control of a split-type drill pipe box robotic arm according to claim 2, characterized in that, The process of using the YOLO target detection algorithm to obtain the distribution status of drill pipes in each compartment includes: The acquired panoramic image is divided into The grid; After predicting the probabilities of B bounding boxes and their classes using each grid cell, the results are integrated and output as a single value. tensor; By using non-maximum suppression processing, overlapping low-confidence prediction boxes are removed, and the final output is the location information of all grid cells that have been identified as drill pipes.

4. The method for inspection and control of a split-type drill pipe box manipulator according to claim 1, characterized in that, After the robotic arm moves to the predetermined position of the target compartment according to the speed curve, it uses a local sensing unit installed on the robotic arm body to perform local detection of the target compartment, and obtains the precise position information and quantity status information of the drill rods in the target compartment, including: The robotic arm moves to the theoretical position above the target compartment according to the speed curve; Local images of the target compartment are captured by an industrial camera mounted on a robotic arm; A vision algorithm based on geometric matching is used to identify the actual center coordinates of the drill pipe in the image and compare them with the theoretical center coordinates of the grid to calculate the positional deviation. The robot arm is controlled to perform offset compensation based on the position deviation, so that the center of the robot arm end effector is aligned with the center of the drill pipe; Then, the laser rangefinder is activated to measure the surface profile of the drill rods in the compartment. Based on the height change steps, the number of drill rod layers and quantities in the current compartment are calculated and updated to the drill rod quantity status information.

5. The method for inspection and control of a split-type drill pipe box manipulator according to claim 4, characterized in that, The method employs a geometric matching-based visual algorithm to identify the actual center coordinates of the drill pipe in the image and compares them with the theoretical center coordinates of the storage grid to calculate the positional deviation, including: After calibrating the camera and hand-eye coordination, standard drill pipe images were acquired. Select a region of interest containing stable features in the standard drill pipe image, and create a geometric matching template based on the region of interest; The central controller triggers the camera to acquire images of the target cell. After preprocessing the acquired target grid image, template matching is performed to calculate the similarity score between the template and the image edge, and the best matching position is obtained. The algorithm outputs matching position transformation parameters, which include the image coordinates and rotation angle of the drill pipe center; The central controller calculates the position deviation based on the transformation relationship between the coordinates and the robot's coordinate system.

6. The method for inspection and control of a split-type drill pipe box manipulator according to claim 4, characterized in that, The formulas for calculating the number and quantity of drill pipe layers in the current compartment based on the height change steps are as follows: in, This refers to the number of drill pipe layers. The equivalent optical height difference for a single-layer drill pipe. This represents the average height of the entire effective contour region. This is the original distance value when the warehouse is empty. This is the distance value below one standard drill pipe layer. Let X be the coverage width of a certain platform in the X direction. The diameter of the drill pipe. The gap between adjacent drill pipes. This refers to the number of drill pipes in a single layer.

7. The method for inspection and control of a split-type drill pipe box manipulator according to claim 1, characterized in that, Based on the obtained precise position and quantity information of the drill rod, the robot arm is controlled to adjust its end effector posture and perform pick-and-place operations, including: Based on the confirmed number of drill pipes, the robotic arm plans the vertical descent speed curve and descends into the target compartment; During the descent, the proximity switch signal installed at the end of the robotic arm and the vertical displacement sensor signal are monitored in real time. If a valid signal from the proximity switch is received within the expected height range, the drill pipe is determined to be present, and the gripper is controlled to perform a gripping action. If no valid proximity switch signal is received after descending to the preset limit position, the compartment is determined to be empty, the robotic arm is raised and switched to the next target compartment according to the decision logic; If abnormal pressure or no signal from the proximity switch is detected during descent or gripping, the robot arm's operation will be stopped immediately and an alarm will be triggered.

Images