An Image-Based Intelligent Control Method for Bucket Wheel Excavators

By using image recognition and iterative fine-tuning feed technology, the problem of inaccurate judgment of the first cut position of the bucket wheel stacker-reclaimer after the coal pile shape changes has been solved, and a more efficient and stable material reclaiming process has been achieved.

CN123105310APending Publication Date: 2026-11-17BEIJING ZHONGSHENG BOFANG ENVIRONMENTAL PROTECTION ENG TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611274510.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

When an existing bucket wheel stacker-reclaimer is restarted after a long period of shutdown, the change in the shape of the coal pile can lead to inaccurate judgment of the first cutter's entry position, which can easily cause empty cuts, shallow cuts, and load impacts, reducing the efficiency of material reclaiming.

Method used

By recognizing the physical contact state between the bucket wheel and the coal pile through image recognition, the disturbance of surface floating coal and the effective cutting state are distinguished. Iterative fine-tuning of the feed is used to determine the actual cutting position, and the bucket wheel attitude is adjusted according to the first cut entry correction amount to perform the first cut material collection.

Benefits of technology

It improves the accuracy of the first cutter's material removal position and the stability of the material removal process, reduces the risks of air cutting, shallow cutting and load impact, and enhances the adaptability of the bucket wheel stacker-reclaimer to changes in the state of the coal pile.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123105310A_ABST
    Figure CN123105310A_ABST
Patent Text Reader

Abstract

This invention discloses an image-based intelligent control method for bucket wheel excavators, belonging to the field of automatic control technology for bucket wheel stacker-reclaimers. Addressing the problem of discrepancies between the predicted and actual cutting positions caused by coal pile settling and surface floating coal after shutdown, this invention, after determining the predicted cutting position, controls the bucket wheel to approach the coal pile at a trial speed. Based on continuous image recognition, it identifies the physical contact state between the bucket wheel and the coal pile, further distinguishing between surface floating coal disturbance and effective cutting states. When in the surface floating coal disturbance state, the bucket wheel is controlled to finely adjust its feed step by step, and continuous images are used to determine whether a continuous coal flow has formed behind the bucket teeth until an effective cutting state is determined. The corresponding bucket wheel position is determined as the actual effective cutting position, and the first cut correction amount is determined based on its difference from the predicted cutting position. After adjusting the bucket wheel attitude, the first cut is executed, thereby improving the accuracy and stability of the first cut.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of automatic control technology for bucket wheel stacker-reclaimers, and in particular to an image-based intelligent control method for bucket wheel stackers. Background Technology

[0002] Bucket wheel stacker-reclaimers are core equipment used for stacking and retrieving materials such as coal in bulk material yards of thermal power plants and ports. To improve operational efficiency and automation, modern bucket wheel stacker-reclaimers are typically equipped with PLC control systems and drive mechanisms for the trolley's travel, rotation, and pitch, as well as sensing devices such as 3D laser scanners to achieve automated control of the material reclaiming operation.

[0003] In existing automated material handling processes, the control system typically acquires the three-dimensional shape of the coal pile using a 3D laser scanning device and establishes a coal pile elevation model. Then, based on the material handling task and the coal pile model, it plans the material handling path and determines the theoretical position where the bucket wheel first cuts into the coal pile, i.e., the first cut position. Finally, the control system drives the trolley traveling mechanism, slewing mechanism, and pitching mechanism of the bucket wheel stacker-reclaimer to move the bucket wheel to the preset position and perform the first cut material handling.

[0004] However, coal piles are bulk material accumulations, and their surface morphology is prone to change over time. Especially when a bucket wheel stacker-reclaimer stops for a long period of time after completing a cycle of operation and then restarts, the coal pile may experience natural settlement, slope slippage, or local collapse due to its own weight, changes in ambient temperature and humidity, etc., causing the three-dimensional model of the coal pile established before the shutdown to deviate from the actual surface morphology of the coal pile at the time of restart.

[0005] Furthermore, the surface of coal piles typically contains loose, loose coal with low physical compaction. When the bucket wheel approaches the coal pile according to the predicted cutting position determined by existing models, the bucket teeth may only contact and agitate the surface loose coal without actually cutting into the coal seam with stable cutting capabilities. Existing control systems usually determine whether cutting has been completed based on changes in the bucket wheel drive current or whether the bucket wheel has reached the preset position. However, in the presence of surface loose coal, it is difficult to accurately determine whether the bucket wheel has entered an effective cutting state based solely on the above information.

[0006] In the aforementioned scenarios, if the system misjudges that the bucket wheel has effectively cut in and directly enters the normal material handling process, the bucket wheel may further cut into the lower coal seam after clearing away the surface coal, causing a sudden increase in drive load, triggering a significant load impact, and even leading to a protective shutdown of the equipment. If the bucket wheel continues to make shallow cuts or empty cuts in the surface coal area, it may result in insufficient material handling, no material in the belt conveyor system, or unstable material flow, reducing material handling efficiency. Therefore, existing technologies have shortcomings in dealing with the first cut operation scenario after a long period of shutdown, including the inability to accurately determine the actual effective cutting position, the susceptibility to load impact, and low material handling efficiency. Summary of the Invention

[0007] To address this, the present invention provides an image-based intelligent control method for bucket wheel excavators. This method overcomes the problems in the prior art where the predicted cutting position is inconsistent with the actual effective cutting position due to changes in the coal pile shape and surface floating coal after shutdown, leading to inaccurate judgment of the first cutter's cutting position and resulting in empty cuts, shallow cuts, and load impacts. This is achieved by identifying the actual contact state between the bucket wheel and the coal pile and determining the actual effective cutting position based on the continuous coal flow.

[0008] To achieve the above objectives, this invention provides an image-based intelligent control method for bucket wheel excavators, applied to bucket wheel stacker-reclaimers. The bucket wheel stacker-reclaimer is equipped with a camera for acquiring images of the coal pile, comprising: Determine the predicted cutting area and predicted cutting position corresponding to the bucket wheel stacker-reclaimer; The bucket wheel stacker-reclaimer is controlled to approach the predicted cutting area at a trial speed lower than the normal operating speed; Based on the images captured by the camera, the physical contact state between the bucket wheel and the coal pile is identified; After identifying the physical contact state, it is determined based on image analysis whether the bucket wheel is in a state of surface coal disturbance. The state of surface coal disturbance indicates that the bucket wheel has not formed an effective cutting after contacting the coal pile. When it is determined that the surface coal is in a disturbed state, the bucket wheel stacker-reclaimer is controlled to perform iterative fine-tuning feed, and after each fine-tuning feed, it is determined whether the bucket wheel has entered an effective cutting state based on the image captured by the camera. When it is determined that the effective cutting state has been entered, the corresponding bucket wheel posture is determined as the actual effective cutting position, and the first cut correction amount is determined based on the predicted cutting position and the actual effective cutting position. Based on the first cut correction amount, the bucket wheel stacker-reclaimer is controlled to adjust its posture and perform the first cut material reclaim.

[0009] Further, determining the predicted cutting area and predicted cutting position corresponding to the bucket wheel stacker-reclaimer includes: Acquire at least one of the following: three-dimensional scanning data of the coal pile, attitude data of the bucket wheel stacker-reclaimer, and historical material reclaiming records; The current coal surface area is determined based on the three-dimensional scanning data, and the corresponding material handling area is determined by combining the attitude data of the bucket wheel stacker-reclaimer and / or the historical material handling records. The predicted cutting area is determined based on the current coal face area and the material handling area, and the position within the predicted cutting area corresponding to the target cutting position of the bucket wheel is determined as the predicted cutting position.

[0010] Furthermore, identifying the physical contact state between the bucket wheel and the coal pile includes: By analyzing the continuous images captured by the camera, when at least one of the following is detected—a change in the contour of the coal surface in front of the bucket wheel, movement of coal particles, or localized breakage—it is determined that the bucket wheel is in physical contact with the coal pile.

[0011] Furthermore, the iterative fine-tuning feed includes: The cantilever of the bucket wheel stacker-reclaimer is controlled to descend gradually in a preset small step length along the direction of entering the coal pile, and the current bucket wheel posture is maintained after each descent of the preset small step length in order to re-acquire the image of the bucket wheel's working area; When the re-acquired image corresponds to the surface coal disturbance state and it is not determined that the bucket wheel has entered the effective cutting state, the boom is controlled to continue descending by the next preset small step.

[0012] Further, determining whether the bucket wheel has entered an effective cutting state includes: The movement area of ​​coal particles behind the bucket wheel teeth is detected based on continuous images acquired by the camera. When the coal particle movement area extends backward from the bucket tooth action area along the rotation direction of the bucket wheel and persists in multiple consecutive sampling frames, it is determined that the bucket wheel has entered the effective cutting state.

[0013] Furthermore, before identifying the physical contact state between the bucket wheel and the coal pile, the method further includes: Obtain the environmental interference status corresponding to the images captured by the camera; When the environmental interference state includes at least one of dust interference, water vapor interference, or low light interference, the corresponding image enhancement processing is performed on the image captured by the camera. Based on the image enhancement processing, the physical contact state, the surface coal disturbance state, and the effective cutting state are identified.

[0014] Furthermore, after performing the first cut and material removal, the process also includes: Acquire images of the coal pile, bucket wheel drive current, and belt scale flow data during the first material handling process; The effective cutting state is determined based on the coal pile image; the bucket wheel drive current is used to determine whether the bucket wheel material handling load has changed abnormally; and the belt scale flow rate data is used to determine whether the material handling flow rate is continuously output. When the effective cutting state continues, the bucket wheel material handling load does not change abnormally, and the material handling flow rate continues to be output, it is determined that the first cut material handling process is in a stable material handling state.

[0015] Furthermore, after determining that the first cutting process is in a stable cutting state, the process further includes: Control the bucket wheel stacker-reclaimer to enter the continuous material reclaiming stage; During the continuous material handling phase, the coal pile image, the bucket wheel drive current, and the belt scale flow data are continuously acquired. When at least one of the following conditions is detected based on the coal pile image: coal flow position deviation, local coal surface collapse, or coal pile boundary change, the cantilever pitch position, slewing position, and / or trolley feed position of the bucket wheel stacker-reclaimer are adjusted to keep the bucket wheel in an effective material-reclaiming state.

[0016] Furthermore, the method of the present invention further includes: Save the predicted cutting position, the actual effective cutting position, and the first cutting correction amount corresponding to the first cut material pick-up; When a subsequent material handling task corresponds to a work area where the first cutting position has been saved, the predicted cutting position corresponding to the subsequent material handling task is corrected based on the positional difference between the saved actual effective cutting position and the predicted cutting position.

[0017] On the other hand, the present invention also provides an image-based intelligent control device for bucket wheel excavators, including a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the image-based intelligent control method for bucket wheel excavators described in the present invention.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: Specifically, the present invention continuously identifies the physical contact state between the bucket wheel and the coal pile, the disturbance state of the surface floating coal, and the effective cutting state through image recognition. When a continuous coal flow is not formed, the actual effective cutting position is gradually determined by iterative fine-tuning of the feed. Then, the first cut correction amount is determined based on the difference between the actual effective cutting position and the predicted cutting position. This allows the first cut to be based on the actual coal surface cutting state to determine the cutting position, reducing the risk of empty cutting, shallow cutting, and load impact caused by changes in the surface floating coal and coal pile morphology, and improving the accuracy of the first cut position and the stability of the material handling process. At the same time, through stability verification after the first cut, dynamic correction during the continuous material handling stage, and the utilization of historical first cut correction data, the adaptability of the bucket wheel stacker reclaimer to changes in the coal pile state is improved. Attached Figure Description

[0019] Figure 1 This is an overall flowchart of the image-based intelligent control method for bucket wheel excavators according to an embodiment of the present invention.

[0020] Figure 2 This is a detailed flowchart of the iterative fine-tuning of feed and effective cutting determination in an embodiment of the present invention.

[0021] Figure 3This is a flowchart illustrating the determination of the stable state of the first material intake and the transition to the continuous material intake stage in an embodiment of the present invention.

[0022] Figure 4 This is a flowchart illustrating adaptive correction based on historical first-cut data in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below with reference to specific embodiments. It should be noted that the following specific embodiments are used to illustrate the technical solutions of this invention, and not to limit the scope of protection of this invention.

[0024] This invention provides an intelligent control method and device for the first cut of a bucket wheel stacker-reclaimer, applicable to scenarios where the bucket wheel stacker-reclaimer restarts material collection after a shutdown. In this scenario, the coal pile undergoes natural settling and redistribution of surface coal during shutdown, and the actual coal surface condition may differ from the 3D model established before shutdown or historical material collection data. This causes the predicted cutting position determined based on existing data to be inconsistent with the actual effective cutting position of the bucket wheel. If the cutting position is determined solely based on the bucket wheel contacting the coal surface or detecting disturbance on the coal surface, the bucket wheel may first contact the loose surface coal without entering the coal seam capable of forming a stable coal flow. Continuing to operate according to normal material collection parameters can easily result in insufficient actual cutting depth or significant load changes during subsequent cutting processes.

[0025] To address the above issues, this invention identifies the contact state between the bucket wheel and the coal surface using continuous images after the bucket wheel approaches the predicted cutting position, further distinguishing between the surface coal disturbance state and the effective cutting state. When it is determined that the bucket wheel only causes surface coal disturbance and has not yet formed a continuous coal flow, normal material handling is not initiated directly. Instead, the bucket wheel stacker-reclaimer is controlled to perform iterative fine-tuning feed, gradually cutting the bucket wheel into the coal seam. After each fine-tuning feed, the movement state of the coal particles behind the bucket teeth is assessed using continuous images, with the formation of a continuous coal flow serving as the criterion for entering the effective cutting state. After determining the bucket wheel posture corresponding to the effective cutting state, this posture is taken as the actual effective cutting position and compared with the predicted cutting position to determine the first cut correction amount. Then, the posture of the bucket wheel stacker-reclaimer is adjusted according to the first cut correction amount before performing the first cut material handling.

[0026] Through the above processing, the initial contact of the coal surface, disturbance of surface floating coal, gradual cutting and formation of effective cutting are continuously judged during the first cutting process. The predicted cutting position is corrected by using the actual effective cutting position, so that the cutting position of subsequent first cutting is determined by the actual coal surface condition, rather than just by the model or historical data before shutdown.

[0027] To facilitate understanding of the technical solution of this invention, the terminology involved in this invention will be explained below: The effective cutting state is defined as follows: after the bucket wheel teeth cut into the coal seam, the coal particles cut by the bucket teeth form a material movement area with continuous spatial extension and continuous movement behind the bucket teeth during the rotation of the bucket wheel, and can continuously generate a coal flow. The iterative fine-tuning feed is defined as follows: under the condition of surface floating coal disturbance, the bucket wheel stacker-reclaimer is controlled to feed towards the coal seam step by step according to a preset small step size, and an image is collected after each feed to determine the state. When a continuous coal flow has not yet been formed, the next fine-tuning feed is performed; when a continuous coal flow is formed, the fine-tuning feed is stopped. The actual effective cutting position is the bucket wheel attitude position corresponding to the first detection of a continuous coal flow during the iterative fine-tuning feed process. The bucket wheel attitude position includes at least the pitch position of the bucket wheel stacker-reclaimer, and may include the rotation position and the trolley position. The first cut correction amount is the difference between the predicted cut position and the actual cut position. The positional difference between the actual effective cutting positions is used to correct the target position and / or attitude parameters corresponding to the first cut material collection, so that the control position of the first cut material collection corresponds to the actual effective cutting position. The first cut cutting correction amount includes at least one of pitch adjustment amount, slewing adjustment amount, and trolley position adjustment amount. The trial speed is the operating speed of the bucket wheel stacker-reclaimer when it approaches the predicted cutting area, which is lower than the normal material collection operation speed. It is used to reduce the feed speed and bucket wheel cutting speed when the bucket wheel first contacts the coal surface. The trial speed can be obtained by reducing at least one of the bucket wheel speed, cantilever pitch speed, trolley travel speed, and slewing speed. The stable material collection state is that during the first cut material collection process, a continuous coal flow can be continuously formed, the bucket wheel drive current does not show abnormal changes such as continuous increase or exceeding the preset allowable fluctuation range, and the belt scale can continuously detect the material collection flow rate within the preset flow range.

[0028] To implement the above method, the bucket wheel stacker-reclaimer system used in this invention includes an image acquisition module, an attitude acquisition module, a material handling status acquisition module, and a control module. The image acquisition module is used to acquire images of the coal surface in front of the bucket wheel. It can employ an industrial camera installed at the front end of the bucket wheel stacker-reclaimer's cantilever, and is equipped with supplementary lighting devices, such as 850nm near-infrared lamps, to improve image clarity of the coal surface and coal particle movement area under low-light conditions. The system may also include a laser scanning module for acquiring three-dimensional scanning data of the coal pile surface. The attitude acquisition module is used to acquire attitude data such as the trolley position, cantilever rotation angle, and cantilever pitch angle of the bucket wheel stacker-reclaimer, which can be obtained through encoders, position sensors, or positioning systems. The material handling status acquisition module is used to acquire operating condition data during the material handling process, such as the bucket wheel drive current and belt scale flow rate. The control module receives the data acquired by the above modules and sends control commands to the actuators of the bucket wheel stacker-reclaimer based on the processing results. The control module can be implemented using an industrial computer or an edge computing host combined with a PLC.

[0029] Specifically, this invention establishes a data acquisition and control pathway for the first cut material reclaiming process of a bucket wheel stacker-reclaimer through the aforementioned modules. The image acquisition module acquires continuous images of the bucket wheel's tooth action area and the coal surface behind it; the attitude acquisition module acquires the bucket wheel's attitude data corresponding to the current image; and the material reclaiming status acquisition module acquires the bucket wheel's drive current and belt scale flow rate during the first cut material reclaiming process. When the industrial computer receives the above data, it determines whether the bucket wheel is in contact with the coal surface based on the continuous images, and further determines whether it is currently in a state of surface coal disturbance or an effective cutting state. After determining the actual effective cutting position, it determines the first cut entry correction amount based on the corresponding bucket wheel attitude data and sends the control command corresponding to the first cut entry correction amount to the PLC. The PLC controls the corresponding actuators in the trolley traveling mechanism, cantilever slewing mechanism, and cantilever pitching mechanism to adjust the bucket wheel attitude according to the control command, thereby enabling image acquisition, status judgment, position correction, and bucket wheel attitude adjustment to be executed sequentially according to the aforementioned method steps.

[0030] During data transmission, image data, laser scan data, attitude data, and material handling status data can be sent to the industrial computer according to their respective sampling periods. The industrial computer selects the corresponding data for processing based on the current first cut material handling stage: during the process of approaching the predicted cutting area, attitude data and image data are mainly used to determine the contact state between the bucket wheel and the coal surface; after identifying the disturbance state of surface floating coal, the images captured by the camera are continuously processed to determine whether a continuous coal flow has formed behind the bucket teeth; after determining the actual effective cutting position, the first cut correction amount is determined based on the attitude data corresponding to the current position and the attitude data corresponding to the predicted cutting position; after completing the first cut material handling, the first cut material handling status is verified by combining the image, bucket wheel drive current, and belt scale flow rate.

[0031] Specifically, this invention establishes a correspondence between changes in the coal surface in the image and the actual posture of the bucket wheel excavator. This ensures that each image state judgment corresponds to the actual position of the current bucket wheel, rather than judging whether normal material collection can proceed based solely on a single coal surface disturbance. Therefore, after identifying surface coal disturbance, the bucket wheel can be fine-tuned based on changes in the coal particle movement area in subsequent images. Furthermore, the bucket wheel posture is recorded when a continuous coal flow is confirmed, providing a positional basis for determining the initial cutter entry correction.

[0032] Please see Figure 1 As shown, the intelligent control method for the first cutter of a bucket wheel stacker-reclaimer according to the present invention includes the following steps.

[0033] Upon receiving the initial material handling task, the process S101, which determines the predicted cutting area and predicted cutting position, is executed. Based on the 3D scanning data of the coal pile, the attitude data of the bucket wheel stacker-reclaimer, and / or historical material handling records, the area where the bucket wheel is expected to enter the coal pile is determined, and the predicted cutting position is determined within the area, serving as a reference position for the subsequent approach and cutting of the bucket wheel.

[0034] After determining the predicted cutting position, the bucket wheel stacker-reclaimer is controlled to approach the predicted cutting area S102 at a trial speed. The trial speed is lower than the normal operating speed of the bucket wheel stacker-reclaimer during material handling, so that the bucket wheel gradually approaches the predicted cutting area, and images of the coal surface in front of the bucket wheel are continuously acquired during the approach.

[0035] Based on the coal surface image, the physical contact state between the bucket wheel and the coal pile is identified in step S103. When a change in the contour of the coal surface, movement of coal particles, or local breakage is detected in the continuous image corresponding to the area of ​​action of the bucket teeth, it is determined that the bucket wheel and the coal pile are in physical contact, and the physical contact state is output.

[0036] After determining the physical contact state, the surface floating coal disturbance state identification S104 is performed. Based on the continuous image corresponding to the physical contact state, it is determined whether the material disturbance in the bucket tooth action area is limited to the vicinity of the bucket tooth, and whether a continuous coal flow is formed behind the bucket tooth; when the material disturbance is limited to the vicinity of the bucket tooth and no continuous coal flow is formed, it is determined that the current state is surface floating coal disturbance state, and the surface floating coal disturbance state is output.

[0037] After determining the surface coal disturbance state, iterative fine-tuning feed S105 is executed. Based on the surface coal disturbance state, the bucket wheel stacker-reclaimer is controlled to perform a fine-tuning feed along the direction of cutting into the coal pile, and after completing the current fine-tuning feed, the current position of the bucket wheel is maintained to acquire corresponding continuous images.

[0038] Based on the continuous images after the fine-tuned feed, the effective cutting state judgment S106 is executed. The coal particle movement area behind the bucket teeth is detected, and it is determined whether the coal particle movement area extends continuously from the bucket tooth action area to the rear of the bucket wheel rotation direction, and whether the extension state persists in the continuous sampling frames; when the judgment condition for forming a continuous coal flow is met, it is determined that the bucket wheel has entered the effective cutting state, and the position corresponding to the current bucket wheel posture is determined as the actual effective cutting position; when the judgment condition is not met, the iterative fine-tuning feed S105 is returned, the bucket wheel is controlled to perform fine-tuning feed again, and the effective cutting state judgment S106 is re-executed.

[0039] After determining the actual effective cutting position, the first cutter entry correction amount determination S107 is executed. The bucket wheel attitude data corresponding to the predicted entry position and the actual effective cutting position are acquired, and the first cutter entry correction amount is determined based on the positional and attitude differences between the two. The first cutter entry correction amount is used to determine the control position and material handling attitude parameters corresponding to the first cutter material handling.

[0040] Based on the first cutter entry correction amount, the first cutter material removal S108 is executed. According to the control position corresponding to the predicted entry position and the first cutter entry correction amount, the control position and material removal posture parameters corresponding to the first cutter material removal are determined, and the bucket wheel stacker reclaimer is controlled to perform the first cutter material removal according to the determined control position and material removal posture parameters, so that the control position corresponding to the first cutter material removal corresponds to the actual effective cutting position.

[0041] Specifically, this invention sets up a continuous image judgment and fine-tuning feed process between the predicted cutting position and the actual effective cutting position, so that after the bucket wheel first contacts the coal pile, it will not directly enter the normal material collection based on the contact event, but will determine the actual effective cutting position based on whether a continuous coal flow is formed behind the bucket teeth, and use the actual effective cutting position to correct the predicted cutting position, so that the bucket wheel posture used for subsequent first cut material collection corresponds to the actual coal surface cutting state.

[0042] Based on the above scheme, the determination of the predicted cutting area and predicted cutting position S101 is explained in detail. In this embodiment, the system acquires the three-dimensional scanning data of the coal pile to be retrieved, the attitude data of the bucket wheel stacker-reclaimer, and historical material retrieval records. Specifically, the three-dimensional scanning data of the coal pile surface is acquired through the laser scanning module to determine the current spatial position of the coal surface; the attitude acquisition module of the bucket wheel stacker-reclaimer acquires the position of the trolley, the cantilever rotation angle, and the cantilever pitch angle, and determines the current position of the bucket wheel in the corresponding coordinate system based on the attitude data; simultaneously, the material retrieval trajectory and corresponding material retrieval area in the historical material retrieval records are acquired.

[0043] The system determines the working area corresponding to the current coal face based on the 3D scanning data, and correlates the working area with the historical material handling trajectory and the current position of the bucket wheel to determine the predicted cutting area when the bucket wheel performs the first material handling operation. Subsequently, based on the coal face position and the preset material handling position within the predicted cutting area, a reference position within the predicted cutting area is determined as the predicted cutting position, and the bucket wheel attitude data corresponding to the predicted cutting position is recorded. The bucket wheel attitude data includes at least one of the following: trolley position, cantilever rotation angle, and cantilever pitch angle.

[0044] Specifically, this invention establishes a correspondence between the current coal face position, the current bucket wheel position, and the historical material collection trajectory, giving the predicted cutting position a clear spatial basis. This predicted cutting position is then used as the starting reference position for subsequent trial approach and image detection. The predicted cutting position is subsequently corrected using the actual effective cutting position, ensuring that the difference between the predicted position and the actual coal face cutting state is recorded and utilized during the first material collection operation.

[0045] Furthermore, the identification of the physical contact state between the bucket wheel and the coal pile, S103, is described in detail. The system continuously acquires images of the coal surface area in front of the bucket wheel teeth using a camera installed on the cantilever of the bucket wheel stacker-reclaimer, and processes the continuous image frames according to a preset image sampling period. The image processing area includes the coal surface area that the bucket teeth are about to enter and the coal surface area near the action area of ​​the bucket teeth.

[0046] During physical contact recognition, the system compares images from adjacent sampling times to detect whether there are image changes in the coal surface area in front of the bucket teeth that correspond to the approaching action of the bucket wheel. When at least one of the following is detected: local displacement or deformation of the coal surface contour, rolling or sliding of coal particles, or local compaction or breakage of the coal surface texture, and the image changes correspond in time to the approaching process of the bucket wheel, it is determined that the bucket wheel has made physical contact with the coal pile, and the corresponding physical contact state is output. As a specific implementation method, inter-frame difference, background subtraction, or motion region detection methods can be used to identify the above image changes.

[0047] After the physical contact state is determined, the system continues to analyze subsequent continuous images instead of directly using the physical contact state as the basis for judging whether to enter the normal material handling state, so as to further determine whether the current contact only occurs in the floating coal area on the surface of the coal pile.

[0048] Specifically, this invention detects changes in continuous images of the bucket tooth's action area, identifying contact features such as coal surface contours, coal particle movement, or localized breakage before a significant change in the bucket wheel's driving load, thereby obtaining the time point of the bucket wheel's first contact with the coal surface. This physical contact state serves as input for subsequent judgment of surface coal disturbance, ensuring that surface coal identification is based on actual contact events, rather than directly determining whether the bucket wheel has entered the coal seam based on a preset position.

[0049] As a specific implementation method, the specific execution logic of the iterative fine-tuning feed S105 is as follows: Figure 2 As shown. Once the system determines from continuous images that it is currently in a state of surface coal disturbance, the iterative fine-tuning feed process is initiated. The control module sends a fine-tuning feed command to the PLC, controlling the cantilever of the bucket wheel stacker-reclaimer to descend gradually in preset small steps along the direction in which the bucket wheel further cuts into the coal pile, such as controlling the cantilever to descend by the preset small step S201. The preset small step can be preset according to the control accuracy of the bucket wheel stacker-reclaimer, the thickness of the surface coal in the coal pile, and the on-site reclaiming conditions, for example, set to 10 mm to 30 mm, and in this embodiment, it is set to 20 mm.

[0050] After completing one fine-tuning feed, the PLC controls the boom to stop descending and maintain the current bucket wheel posture, so that the camera can stably acquire continuous images of the bucket wheel tooth action area and the coal surface behind it (S202). The control module analyzes the re-acquired continuous images to determine whether a continuous coal flow has formed behind the bucket teeth (S203). When the continuous images indicate that a continuous coal flow has formed behind the bucket teeth, it is determined that the bucket wheel has entered an effective cutting state, the fine-tuning feed is stopped, and the bucket wheel posture data corresponding to the current moment is acquired. The position corresponding to the bucket wheel posture data is determined as the actual effective cutting position (S204).

[0051] When the continuous images indicate that no continuous coal flow has been formed behind the bucket teeth, it is determined that the bucket wheel has not yet entered the effective cutting state and is still in the state of surface floating coal disturbance. Then, return to step S201 and control the boom to continue to descend by the next preset small step S205. After completing the next descent, maintain the current bucket wheel posture again and collect continuous images for judgment.

[0052] The aforementioned fine-tuning of the feed and image judgment are repeated until a continuous coal flow is determined to be formed behind the bucket teeth based on the continuous images. When a continuous coal flow is determined to be formed, the descent is stopped, and the bucket wheel attitude data corresponding to the current moment is acquired. The position corresponding to the bucket wheel attitude data is determined as the actual effective cutting position.

[0053] Specifically, after identifying surface coal disturbance, this invention does not directly control the bucket wheel to enter the preset normal cutting depth. Instead, it gradually changes the bucket wheel's entry position in small steps, and re-acquires the coal surface image after each change. Thus, the subsequent entry depth is determined by the result of the previous image assessment, enabling the bucket wheel to gradually find an effective cutting position between the surface coal and the actual cutable coal seam, avoiding a large change in entry depth at once that would cause the bucket wheel to directly enter an unconfirmed coal seam.

[0054] Furthermore, the process for determining the effective cutting state S106 is explained. After each fine-tuning feed is completed, the control module acquires continuously captured images of the coal surface behind the bucket wheel teeth from the camera, and determines the coal particle movement area from the images. The coal particle movement area is the material region in the continuous images whose position changes relative to the coal surface background. As a specific implementation, the coal particle movement area can be obtained using optical flow, inter-frame difference, or motion region detection methods.

[0055] After determining the coal particle movement area, the control module further judges the positional change of the coal particle movement area relative to the bucket tooth action area. When the coal particle movement is only concentrated near the bucket teeth, and the movement area does not form a continuous extension area behind the bucket wheel in the direction of rotation, it is determined that a continuous coal flow has not yet formed. When the coal particle movement area extends from the bucket tooth action area to the rear of the bucket wheel in the direction of rotation, and forms a continuous movement area relative to the direction of bucket wheel movement, subsequent sampling frames are further judged.

[0056] Specifically, multiple sampling frames can be acquired continuously, and the coal particle movement area in each sampling frame can be determined. When the coal particle movement area maintains a state of extending backward from the bucket tooth action area in multiple consecutive sampling frames, and the movement area in adjacent sampling frames remains continuous in spatial position, it is determined that a continuous coal flow has formed behind the bucket tooth, and the current state is determined as an effective cutting state. The number of consecutive sampling frames can be preset according to the camera sampling frequency and the bucket wheel running speed, for example, set to 3 to 5 consecutive frames.

[0057] When the judgment condition for forming a continuous coal flow is not met, the control module outputs the judgment result that it has not entered the effective cutting state and returns to the iterative fine-tuning feed process, so that the cantilever descends again by a small step; when the judgment condition for forming a continuous coal flow is met, the control module outputs the effective cutting state and stops the current iterative feed process.

[0058] Specifically, this invention uses whether the coal particle movement area extends continuously backward from the bucket tooth's action area and whether this extension persists as the image judgment criterion for the effective cutting state, thus giving different image discrimination conditions to surface coal disturbance and the actual cutting process. During surface coal disturbance, coal particle movement is mainly concentrated near the bucket teeth; after entering the effective cutting state, the material carried out by the bucket teeth can form a continuous movement area behind them. Therefore, by confirming this movement area through continuous sampling frames, misjudgments caused by single-frame coal particle slippage or local collapse can be reduced, and image evidence can be provided for determining the actual effective cutting position.

[0059] After determining the first-cut material handling control parameters based on the actual effective cutting position and the first-cut entry correction amount, and executing the first-cut material handling, the system further verifies the stability of the first-cut material handling process. Please refer to [link / reference needed]. Figure 3 As shown.

[0060] First, the corrected first cut material removal S301 is executed. The system adjusts the attitude of the bucket wheel stacker-reclaimer according to the determined first cut infeed correction amount, and performs the first cut material removal according to the adjusted attitude.

[0061] Specifically, this invention corrects the predicted cutting position before the first cut, enabling the bucket wheel to perform the first cut according to the posture corresponding to the actual effective cutting position, thereby matching the cutting position of the first cut with the actual coal face cutting state.

[0062] During the initial material handling process, the system continuously collects operational data S302 corresponding to the material handling process. This operational data includes images of the coal pile, the drive current of the bucket wheel drive mechanism, and the flow rate data of the belt scale corresponding to the belt conveyor mechanism.

[0063] Specifically, this invention acquires multiple types of working condition data reflecting the coal flow status, bucket wheel load status, and material output status during the first cutter material collection process, providing a data basis for subsequent judgment of the first cutter material collection status.

[0064] Subsequently, the system performs a stable material handling status judgment (S303) based on the collected operating condition data. Specifically, the system uses coal pile image detection to determine the coal particle movement area behind the bucket teeth and the continuity of the coal flow to determine whether the bucket wheel is continuously in an effective cutting state; it uses bucket wheel drive current to detect changes in bucket wheel material handling load, and determines that there is an abnormality in bucket wheel material handling load when the drive current continuously increases or the current change exceeds the preset allowable fluctuation range within a preset time period; and it uses belt scale flow rate to detect material output, and determines that material output is normal when the belt scale flow rate is continuously within the preset flow rate range.

[0065] When the effective cutting state continues, the bucket wheel drive current does not change abnormally, and the belt scale flow rate meets the preset flow rate condition, the system determines that the first cut material taking process has entered the stable material taking state S304.

[0066] Specifically, this invention uses coal pile images, bucket wheel drive current, and belt scale flow rate to determine the cutting state, equipment load, and material output during the first cut material handling process, and uses the results of each determination as the basis for determining the stable material handling state, thereby reducing the possibility of misjudgment based on a single operating parameter.

[0067] After determining that the initial material handling process has entered a stable material handling state (S304), the system controls the bucket wheel stacker-reclaimer to enter the continuous material handling stage (S305), and uses the bucket wheel posture and corresponding operating parameters determined during the initial material handling process as the initial control parameters for the continuous material handling stage. The operating parameters include at least one of the bucket wheel rotational speed and the trolley feed speed.

[0068] Specifically, the present invention uses the bucket wheel posture and operating parameters verified by the first cut as the initial control parameters for the continuous feeding stage, so that continuous feeding can continue to be executed based on the verified feeding state, reducing the process of re-determining the initial feeding parameters.

[0069] During the continuous material handling phase, the system continues to collect images of the coal pile, bucket wheel drive current, and belt scale flow data, and performs dynamic correction S306 on the material handling process based on the collected data.

[0070] Specifically, when the coal pile image detects a shift in the movement area of ​​the continuous coal flow, a local collapse of the coal surface, or a change in the relative position between the bucket wheel and the effective material collection area, it is determined that the current material collection position needs to be adjusted; when the bucket wheel drive current continues to rise and exceeds the preset load change condition, it is determined that the bucket wheel cutting load has changed; when the belt scale flow rate is lower than or higher than the preset flow rate range, it is determined that the material output has changed.

[0071] Based on detected changes in the material handling status, the system adjusts at least one of the cantilever pitching mechanism, slewing mechanism, and trolley traveling mechanism. For example, when a gradual weakening of the continuous coal flow and a decrease in the belt scale flow rate are detected, the trolley feed speed is reduced or the cantilever pitching position is adjusted; when a sudden increase in the belt scale flow rate is detected and the coal pile image shows material accumulation in the bucket wheel's action area, the cantilever pitching position or the trolley position is adjusted to change the bucket wheel's material handling position. After adjustment, the system continues to collect coal pile images, bucket wheel drive current, and belt scale flow rate, and reassesses the material handling status based on the adjusted data.

[0072] Specifically, this invention continuously monitors the coal flow status, bucket wheel load status, and material output status during the continuous material feeding stage, and adjusts the corresponding bucket wheel posture or operating parameters according to different status changes. This allows the actual material feeding position of the bucket wheel to be adjusted according to changes in the coal pile status, thereby reducing fluctuations in material feeding flow and abnormal bucket wheel load caused by local changes in the coal pile.

[0073] To correct the predicted cutting position of subsequent material handling tasks based on historical first-cut material handling results, this invention also includes a parameter adjustment process based on historical material handling data, such as... Figure 4 As shown.

[0074] After a first cut material removal task is completed, the first cut material removal task completion step S401 is executed. Once the system confirms that the first cut material removal task has been completed, it records and saves the relevant data from the process S402. The saved data includes the task's corresponding work area identifier, predicted cutting position, actual effective cutting position, and the first cut cutting correction amount determined based on the predicted and actual effective cutting positions. The corresponding task time information can also be saved simultaneously.

[0075] Specifically, this invention saves the correspondence between the predicted cutting position and the actual effective cutting position, so that subsequent material handling tasks can directly call the existing first cut position deviation data, providing a data basis for the correction of the subsequent predicted cutting position.

[0076] When the system receives a new first-cut material handling task, it executes the subsequent material handling task reception S403. The system obtains the working area corresponding to the material handling task and matches the working area with the working areas saved in the historical data to determine whether there is corresponding historical first-cut material handling data in the current working area S404.

[0077] When it is determined that there is corresponding historical first-cut material handling data in the current work area, the system retrieves the first-cut entry correction data S405 corresponding to the current work area from the historical data. The first-cut entry correction data may include the first-cut entry correction amount corresponding to the most recent first-cut material handling task, or it may include the first-cut entry correction amounts corresponding to multiple first-cut material handling tasks within a preset historical time range.

[0078] After acquiring historical first-cut entry correction data, the system adjusts the predicted entry position corresponding to the current material handling task based on this data. For example, when historical data indicates that the actual effective cutting position within the same working area is continuously shifted downwards relative to the predicted entry position, the system determines the corresponding downward correction distance based on the historical first-cut entry correction amount and applies this correction distance to the current predicted entry position to obtain the corrected predicted entry position. Only when the adjustment direction of the historical first-cut entry correction amount remains consistent within the preset historical records is the current predicted entry position pre-corrected based on the historical first-cut entry correction data. Subsequently, the corrected predicted entry position is used as the starting position for the first-cut approach and iterative fine-tuning feed.

[0079] When it is determined that there is no corresponding historical first cut material taking data in the current working area, the original predicted cutting position S406 determined by the current task based on the three-dimensional scanning data, bucket wheel excavator attitude data and / or other material taking data is used, and the original predicted cutting position is used as the starting position for the first cut approach and iterative fine-tuning feed.

[0080] After determining the predicted cutting position corresponding to the current material handling task, the first cut process S407 is executed. Specifically, the system uses the predicted cutting position after correction by the historical first cut or the original predicted cutting position as the starting position of this first cut process, and performs the following steps: approaching the test speed, identifying the physical contact state, judging the disturbance state of the surface floating coal, iteratively fine-tuning the feed, judging the effective cutting state, determining the first cut correction amount, and first cut material handling.

[0081] Specifically, this invention associates the position correction relationship in the historical first cut material handling task with the working area of ​​the current material handling task, and adjusts the current predicted cutting position according to the historical first cut cutting correction data. This allows the working area with historical material handling data to pre-correct the prediction result using the existing position deviation, thereby making the initial position of the subsequent first cut control closer to the historical actual effective cutting position and reducing the number of adjustments required for iterative fine-tuning of the feed.

[0082] To more clearly illustrate the technical solution of the present invention, the first-cutting intelligent control process of the present invention will be described below with reference to a specific embodiment.

[0083] This embodiment employs a bucket wheel stacker-reclaimer equipped with a front-end industrial camera, a bucket wheel drive mechanism, a cantilever pitching mechanism, and a PLC controller. The industrial camera is mounted at the front end of the cantilever of the bucket wheel stacker-reclaimer to acquire continuous images of the area where the bucket wheel teeth act and the coal pile area behind them. The system determines the predicted cutting position corresponding to this task based on historical material reclaiming records, and uses this position as the initial target position for the first cut.

[0084] At the start of this material handling task, the system determines the predicted cutting position and controls the bucket wheel stacker-reclaimer to approach the predicted cutting position at a trial speed lower than the normal material handling speed. In this embodiment, the normal rotational speed of the bucket wheel during material handling is 30-50 rpm. When approaching the predicted cutting position, the bucket wheel rotational speed is set to 15 rpm, and the cantilever pitch descent speed is reduced to decrease the feed speed when the bucket wheel first approaches the coal surface.

[0085] As the bucket wheel approaches the predicted cutting position, an industrial camera continuously captures images of the coal surface area in front of the bucket wheel teeth. The image analysis module compares adjacent sampled frames. When a change in the contour of the coal surface in front of the bucket teeth is detected, and some coal particles are detected rolling or sliding outward from their original positions, the system determines that the bucket wheel has made physical contact with the coal pile.

[0086] After confirming physical contact, the system continues to analyze images of the area affected by the bucket teeth and the area behind them. At this point, it detects that the coal particle movement area is mainly distributed near the bucket teeth, and the movement area does not extend continuously backward along the bucket wheel's rotation direction, nor does a continuous material movement area form behind the bucket teeth. Therefore, the system classifies the current state as surface floating coal disturbance, rather than determining the current position as the actual effective cutting position.

[0087] After confirming that the surface coal is disturbed, the system initiates an iterative fine-tuning feed process. In this embodiment, the single fine-tuning feed step size is set to 20mm, and the cantilever pitching mechanism is controlled to descend sequentially along the direction in which the bucket wheel further cuts into the coal pile. After the first fine-tuning feed is completed, the cantilever descends 20mm and pauses, and the industrial camera re-captures continuous images of the bucket tooth action area and the area behind the bucket teeth. Image analysis reveals that the coal particle movement is still mainly concentrated near the bucket teeth, and no continuous material movement area has been formed behind the bucket teeth. Therefore, the system continues to execute the next 20mm fine-tuning feed.

[0088] After the second fine-tuning feed was completed, the cantilever had descended a cumulative 40mm. The system paused again and acquired continuous images. After analysis, no movement area meeting the criteria for continuous coal flow was detected, so the third fine-tuning feed was executed. After the third fine-tuning feed was completed, the cantilever had descended a cumulative 60mm. The system performed image analysis again, and the results still indicated that the coal particle movement area had not formed a continuous movement area extending along the direction of bucket wheel rotation. Therefore, the fourth fine-tuning feed was executed.

[0089] After the fourth fine-tuning feed, the cantilever has cumulatively descended 80mm relative to its initial fine-tuning position. The system pauses and acquires continuous images. By analyzing the coal particle movement area in the continuous images, a continuous movement area extending backward along the bucket wheel's rotation direction is detected behind the bucket teeth, and this movement area persists across multiple consecutive sampling frames. Based on this, the system determines that the bucket wheel has entered an effective cutting state and identifies the current position corresponding to the bucket wheel's posture as the actual effective cutting position.

[0090] After determining the actual effective cutting position, the system acquires the trolley position, slewing angle, and cantilever pitch angle corresponding to the current bucket wheel, and compares them with the trolley position, slewing angle, and cantilever pitch angle corresponding to the predicted cutting position. For example, if the cantilever pitch angle corresponding to the predicted cutting position is θ1 and the cantilever pitch angle corresponding to the actual effective cutting position is θ2, then the corresponding pitch correction amount is determined based on the angle difference between θ1 and θ2. Similarly, for the trolley position and slewing angle, the corresponding correction amount is determined based on the positional difference between the predicted cutting position and the actual effective cutting position, thus forming the first cut correction amount.

[0091] Subsequently, the controller determines the first cut material handling control position and material handling posture parameters corresponding to the actual effective cutting position based on the control position corresponding to the predicted cutting position and the first cut cutting correction amount. It then controls the corresponding mechanisms in the cantilever pitch mechanism, slewing mechanism, and trolley traveling mechanism to perform the first cut material handling according to the determined control parameters, so that the control position of the formal first cut material handling corresponds to the actual effective cutting position.

[0092] During the initial material handling process, the system continues to collect images of the coal pile, bucket wheel drive current, and belt scale flow rate data. Based on the coal pile image, it determines whether the continuous coal flow persists; based on the bucket wheel drive current, it determines whether there are any abnormal changes in the material handling load; and based on the belt scale flow rate, it determines whether the material output is within the preset flow rate range. When all the above data meet the preset stable material handling conditions, the system determines that the initial material handling process has entered a stable material handling state and transitions to the continuous material handling stage.

[0093] This embodiment uses a bucket wheel speed of 15 rpm, a single fine-tuning feed of 20 mm, and a cumulative fine-tuning of 80 mm as examples to illustrate the complete process from low-speed approach, physical contact identification, judgment of surface coal disturbance, iterative fine-tuning feed, confirmation of effective cutting state, determination of actual effective cutting position, to correction of the first cut. The above parameters are used to illustrate the specific implementation of the present invention. The actual operating parameters can be set according to the structural parameters of the bucket wheel stacker-reclaimer, the state of the coal pile, and the reclaiming conditions.

[0094] Specifically, this invention further determines whether a continuous coal flow has formed in the coal particle movement area after physical contact is detected, and adjusts the bucket wheel cutting position step by step with the feed step size when a continuous coal flow has not formed, so that the actual effective cutting position can be gradually determined through continuous image feedback; after confirming the effective cutting state, the first cut correction amount is determined according to the positional difference between the predicted cutting position and the actual effective cutting position, thereby avoiding directly performing normal material taking when only surface floating coal disturbance is detected.

[0095] Those skilled in the art will understand that the present invention also provides a first-cutting intelligent control device for a bucket wheel stacker-reclaimer. This device is installed in the control system of the bucket wheel stacker-reclaimer and is used to execute the first-cutting intelligent control method. The first-cutting intelligent control device includes a processor and a memory. The memory stores a computer program. When the processor executes the computer program, it causes the processor to perform the following steps: determining the predicted cutting area and predicted cutting position; low-speed approach of the bucket wheel; identification of the physical contact state between the bucket wheel and the coal pile; judgment of the surface floating coal disturbance state; iterative fine-tuning of the feed; judgment of the effective cutting state; determination of the actual effective cutting position; determination of the first-cutting cutting correction amount; and first-cutting material reclaiming control.

[0096] In practical implementation, the first-cutting intelligent control device can be implemented using an industrial computer, edge computing device, or other computing device with data processing and control functions, and is communicatively connected to the camera, attitude acquisition device, bucket wheel drive mechanism, cantilever pitch mechanism, trolley travel mechanism, and slewing mechanism of the bucket wheel stacker-reclaimer. The processor determines the physical contact state between the bucket wheel and the coal pile based on the coal pile image acquired by the camera, and when it is determined that the surface floating coal is disturbed, it controls the cantilever to perform iterative fine-tuning of the feed according to a preset small step size; after detecting that a coal particle movement area that meets the preset continuity condition is formed behind the bucket teeth, the current bucket wheel attitude is determined as the actual effective cutting position, and the first-cutting entry correction amount is determined based on the positional difference between the predicted entry position and the actual effective cutting position, thereby controlling the bucket wheel stacker-reclaimer to adjust its attitude and perform the first-cutting material reclaiming.

[0097] After the initial material removal is completed, the processor can determine whether the initial material removal is in a stable state based on the coal pile image, bucket wheel drive current, and belt scale flow data. Upon entering the continuous material removal phase, it adjusts the trolley position, cantilever pitch position, slewing position, and / or operating speed of the bucket wheel stacker-reclaimer based on detected changes in coal flow, bucket wheel drive current, and belt scale flow. Simultaneously, the processor can associate and store the predicted cutting position, the actual effective cutting position, and the initial cutting correction amount with the corresponding working area, and retrieve the corresponding historical data in subsequent material removal tasks to pre-correct the current predicted cutting position.

[0098] Specifically, this invention enables the first cutter intelligent control device to determine the first cutter entry position based on the material state after the bucket wheel actually contacts the coal pile by performing processes such as predicted position determination, visual contact recognition, fine-tuning feed, effective cutting position determination, and attitude correction in the control device, and to adjust the bucket wheel attitude according to the actual effective cutting position; at the same time, by saving the running data and position correction data after the first cutter picks up material, it can also provide historical correction basis for subsequent material picking tasks.

[0099] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An image-based intelligent control method for a bucket wheel stacker-reclaimer, applied to a bucket wheel stacker-reclaimer equipped with a camera for acquiring images of a coal pile, characterized in that... include: Determine the predicted cutting area and predicted cutting position corresponding to the bucket wheel stacker-reclaimer; The bucket wheel stacker-reclaimer is controlled to approach the predicted cutting area at a trial speed lower than the normal operating speed; Based on the images captured by the camera, the physical contact state between the bucket wheel and the coal pile is identified; After identifying the physical contact state, it is determined based on image analysis whether the bucket wheel is in a state of surface coal disturbance. The state of surface coal disturbance indicates that the bucket wheel has not formed an effective cutting after contacting the coal pile. When it is determined that the surface coal is in a disturbed state, the bucket wheel stacker-reclaimer is controlled to perform iterative fine-tuning feed, and after each fine-tuning feed, it is determined whether the bucket wheel has entered an effective cutting state based on the image captured by the camera. When it is determined that the effective cutting state has been entered, the corresponding bucket wheel posture is determined as the actual effective cutting position, and the first cut correction amount is determined based on the predicted cutting position and the actual effective cutting position. Based on the first cut correction amount, the bucket wheel stacker-reclaimer is controlled to adjust its posture and perform the first cut material reclaim.

2. The image-based intelligent control method for bucket wheel excavators according to claim 1, characterized in that, Determining the predicted cutting area and predicted cutting position corresponding to the bucket wheel stacker-reclaimer includes: Acquire at least one of the following: three-dimensional scanning data of the coal pile, attitude data of the bucket wheel stacker-reclaimer, and historical material reclaiming records; The current coal surface area is determined based on the three-dimensional scanning data, and the corresponding material handling area is determined by combining the attitude data of the bucket wheel stacker-reclaimer and / or the historical material handling records. The predicted cutting area is determined based on the current coal face area and the material handling area, and the position within the predicted cutting area corresponding to the target cutting position of the bucket wheel is determined as the predicted cutting position.

3. The image-based intelligent control method for bucket wheel excavators according to claim 1, characterized in that, The identification of the physical contact state between the bucket wheel and the coal pile includes: By analyzing the continuous images captured by the camera, when at least one of the following is detected—a change in the contour of the coal surface in front of the bucket wheel, movement of coal particles, or localized breakage—it is determined that the bucket wheel is in physical contact with the coal pile.

4. The image-based intelligent control method for bucket wheel excavators according to claim 1, characterized in that, The iterative fine-tuning feed includes: The cantilever of the bucket wheel stacker-reclaimer is controlled to descend gradually in a preset small step length along the direction of entering the coal pile, and the current bucket wheel posture is maintained after each descent of the preset small step length in order to re-acquire the image of the bucket wheel's working area; When the re-acquired image corresponds to the surface coal disturbance state and it is not determined that the bucket wheel has entered the effective cutting state, the boom is controlled to continue descending by the next preset small step.

5. The image-based intelligent control method for bucket wheel excavators according to claim 1, characterized in that, Determining whether the bucket wheel has entered an effective cutting state includes: The movement area of ​​coal particles behind the bucket wheel teeth is detected based on continuous images acquired by the camera. When the coal particle movement area extends backward from the bucket tooth action area along the rotation direction of the bucket wheel and persists in multiple consecutive sampling frames, it is determined that the bucket wheel has entered the effective cutting state.

6. The image-based intelligent control method for bucket wheel excavators according to claim 1, characterized in that, Before identifying the physical contact state between the bucket wheel and the coal pile, the process also includes: Obtain the environmental interference status corresponding to the images captured by the camera; When the environmental interference state includes at least one of dust interference, water vapor interference, or low light interference, the corresponding image enhancement processing is performed on the image captured by the camera. Based on the image enhancement processing, the physical contact state, the surface coal disturbance state, and the effective cutting state are identified.

7. The image-based intelligent control method for bucket wheel excavators according to claim 1, characterized in that, After performing the first cut and material removal, the process also includes: Acquire images of the coal pile, bucket wheel drive current, and belt scale flow data during the first material handling process; The effective cutting state is determined based on the coal pile image; the bucket wheel drive current is used to determine whether the bucket wheel material handling load has changed abnormally; and the belt scale flow rate data is used to determine whether the material handling flow rate is continuously output. When the effective cutting state continues, the bucket wheel material handling load does not change abnormally, and the material handling flow rate continues to be output, it is determined that the first cut material handling process is in a stable material handling state.

8. The image-based intelligent control method for bucket wheel excavators according to claim 7, characterized in that, After determining that the first cutting process is in a stable cutting state, the process further includes: Control the bucket wheel stacker-reclaimer to enter the continuous material reclaiming stage; During the continuous material handling phase, the coal pile image, the bucket wheel drive current, and the belt scale flow data are continuously acquired. When at least one of the following conditions is detected based on the coal pile image: coal flow position deviation, local coal surface collapse, or coal pile boundary change, the cantilever pitch position, slewing position, and / or trolley feed position of the bucket wheel stacker-reclaimer are adjusted to keep the bucket wheel in an effective material-reclaiming state.

9. The image-based intelligent control method for bucket wheel excavators according to claim 1, characterized in that, Also includes: Save the predicted cutting position, the actual effective cutting position, and the first cutting correction amount corresponding to the first cut material pick-up; When a subsequent material handling task corresponds to a work area where the first cutting position has been saved, the predicted cutting position corresponding to the subsequent material handling task is corrected based on the positional difference between the saved actual effective cutting position and the predicted cutting position.

10. An image-based intelligent control device for bucket wheel excavators, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the image-based intelligent control method for bucket wheel excavators as described in any one of claims 1-9.