Alignment control method and system for belt conveyor

Abstract

The invention relates to the technical field of belt conveyor control, discloses an alignment control method and system of a belt conveyor, and aims to solve the problems of poor precision and low efficiency of an existing method. The scheme mainly comprises the steps that a visual detection module is installed below the discharging end of a superior belt conveyor; based on an acquired image of a visual detection module, before equipment is installed and used, an upper-level belt conveyor is adjusted to a target alignment position, and the vertex of the target lower left edge of a balancing weight of a lower-level belt conveyor at the moment is recognized and obtained; in the equipment operation process, the position coordinates of the real-time lower edge vertex of the balancing weight of the lower-stage belt conveyor are recognized and obtained in real time; and the offset is calculated in real time according to the position coordinates of the target left lower edge vertex and the real-time lower edge vertex, and the upper-stage belt conveyor is controlled to conduct position adjustment according to the offset till the upper-stage belt conveyor and the lower-stage belt conveyor are aligned. The grain aligning device improves the aligning precision and efficiency and is particularly suitable for grain conveying.

Description

Technical Field

[0001] This invention relates to the field of belt conveyor control technology, and specifically to an alignment control method and system for belt conveyors. Background Technology

[0002] In the grain storage industry, belt conveyors are the core conveying equipment for grain storage operations. To meet the needs of long-distance and high-drop conveying in grain storage operations, a multi-stage series arrangement is generally adopted, that is, multiple belt conveyors are connected end to end to form a complete conveying line.

[0003] Currently, in the cascaded application of multi-stage belt conveyors, the alignment between two stages—that is, the spatial alignment of the discharge end of the upper conveyor with the feed end of the lower conveyor—relies entirely on manual operation. Operators must visually observe the relative positions of the two stages and, based on experience, manually push the conveyor frame, adjust the anchor bolts, or move the wheels to gradually approach the alignment. This manual alignment method has several significant drawbacks: First, its alignment accuracy depends entirely on the operator's subjective judgment and skill level, making it highly susceptible to deviations due to visual errors, lighting conditions, or fatigue. Second, the manual adjustment process is cumbersome, time-consuming, and labor-intensive, severely impacting the deployment efficiency and automation level of the entire conveying system.

[0004] Due to alignment deviations, material thrown from the upper conveyor belt may not accurately land in the central receiving area of ​​the lower conveyor belt during material conveying, leading to serious material spillage. This not only increases cleanup costs and losses but may also cause malfunctions such as conveyor blockages, belt misalignment, and even drive motor overload damage due to spilled material accumulating around the equipment. Furthermore, slight positional shifts caused by vibration and impact during equipment operation are difficult to detect manually in real time, ensuring a persistent risk of spillage.

[0005] Therefore, the problems of poor alignment accuracy and low efficiency faced by existing two-stage belt conveyors have become key bottlenecks restricting related industries from improving automation levels, ensuring production safety, and optimizing operational efficiency. An intelligent solution capable of achieving automatic and precise alignment is urgently needed. Summary of the Invention

[0006] This invention aims to solve the problems of poor accuracy and low efficiency in the existing alignment methods of two-stage belt conveyors, and proposes an alignment control method and system for belt conveyors.

[0007] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: In a first aspect, the present invention provides an alignment control method for a belt conveyor, the method comprising: Step 1: Install a vision inspection module below the discharge end of the upper belt conveyor. The camera of the vision inspection module is installed parallel to the upper belt conveyor and acquires images that simultaneously include the surface of the upper conveyor belt, the edge contour of the counterweight of the lower belt conveyor, and the ground at the bottom of the lower belt conveyor. Step 2: Before the equipment is installed and used, adjust the upper belt conveyor to the target alignment position. Based on the image acquired by the vision detection module, identify and obtain the position coordinates of the lower left edge vertex B1 and the lower right edge vertex B2 of the target of the counterweight block of the lower belt conveyor at this time. Step 3: During equipment operation, based on the images acquired by the vision detection module, the real-time position coordinates of the lower edge vertex of the counterweight block of the lower belt conveyor are identified and obtained in real time. The real-time lower edge vertex includes the real-time lower left edge vertex A1 and the real-time lower right edge vertex A2. Step 4: Calculate the offset in real time based on the position coordinates of B1, B2, A1 and A2, and control the upper belt conveyor to adjust its position according to the offset until the upper belt conveyor and the lower belt conveyor are aligned.

[0008] Furthermore, in step 3, when only one point between A1 and A2 is identified, the method further includes: Based on the identified X-coordinate misalignment between A1 or A2 and B1, the position of the upper belt conveyor is initially adjusted. After adjustment, if A1 and A2 appear simultaneously in the acquired image, proceed to step 4; if only one of A1 and A2 is in the field of view, adjust the position of the upper belt conveyor again according to the identified X-coordinate misalignment between A1 or A2 and B2, until A1 and A2 appear simultaneously in the acquired image.

[0009] Furthermore, the offset includes rotational offset, horizontal offset, and vertical offset.

[0010] Furthermore, step 4 specifically includes: Calculate the rotational offset, and control the upstream belt conveyor to perform rotational offset correction based on the rotational offset until the rotational offset meets the following requirements: ,in, Indicates the rotation offset. Indicates the rotation offset threshold; Calculate the horizontal offset, and control the upstream belt conveyor to perform horizontal offset correction based on the horizontal offset until the horizontal offset meets the following requirements: ,in, Indicates the horizontal offset. Indicates the horizontal offset threshold; Calculate the longitudinal offset, and control the upstream belt conveyor to perform longitudinal offset correction based on the longitudinal offset until the longitudinal offset meets the following requirements: ,in, Indicates the vertical offset. This indicates the vertical offset threshold.

[0011] Furthermore, the method for calculating the rotational offset includes: Calculate the angle between the line connecting A1 and A2 and the line connecting B1 and B2; this angle is the rotation offset. The calculation formula is as follows: ; in, Indicates the position coordinates of A1. Indicates the position coordinates of A2. Indicates the position coordinates of B1. This represents the position coordinates of B2.

[0012] Furthermore, the horizontal offset The calculation formula is as follows: ; in, The X-coordinate of A1 after rotational offset correction is represented by the following formula: .

[0013] Furthermore, the longitudinal offset The calculation formula is as follows: ; in, The Y-coordinate of A1 after rotational offset correction is represented by the following formula: .

[0014] Furthermore, the image recognition process includes: The acquired images are converted to HSV space and preprocessed using contrast linear transformation. Noise reduction and morphological processing are performed on the images. The Wallner algorithm with adaptive threshold segmentation is used for image binarization segmentation. Connected component contours are extracted, and the edge images of the counterweight blocks are extracted using the Canny edge detection algorithm. Finally, the lower left connection point of the vertical and horizontal lines of the edge images is extracted as A1 or B1, and the lower right connection point is extracted as A2 or B2.

[0015] Furthermore, the method also includes: Images of belt defects, including surface wear, cracks, loose joints, and misalignment, are collected in advance. Machine learning algorithms are then used to train the samples to generate a defect recognition model. The visual inspection module synchronously acquires surface and edge images of the upper conveyor belt; after grayscale conversion and noise reduction preprocessing, the defect recognition model is called to analyze the image and identify whether a preset defect type exists; if a defect is identified, the early warning module is triggered to issue an alarm, and the defect type, location, and severity information are displayed on the display module.

[0016] In a second aspect, the present invention provides an alignment control system for a belt conveyor, used to implement the alignment control method for the belt conveyor described in the first aspect, the system comprising: The vision inspection module is fixedly installed below the discharge end of the upper belt conveyor. The camera of the vision inspection module is installed parallel to the upper belt conveyor. It is used to collect images that simultaneously include the surface of the upper conveyor belt, the edge contour of the counterweight of the lower belt conveyor, and the ground at the bottom of the lower conveyor belt. The control module, communicatively connected to the vision detection module, is used to adjust the upper-level belt conveyor to the target alignment position before the equipment is installed and used. Based on the image acquired by the vision detection module, it identifies and obtains the position coordinates of the lower left edge vertex B1 and the lower right edge vertex B2 of the counterweight block of the lower-level belt conveyor at this time. During equipment operation, based on the image acquired by the vision detection module, it identifies and obtains the real-time position coordinates of the lower edge vertex of the counterweight block of the lower-level belt conveyor, including the real-time lower left edge vertex A1 and the real-time lower right edge vertex A2. The offset is calculated in real time based on the position coordinates of B1, B2, A1, and A2, and corresponding control commands are generated based on the offset. The drive execution module, connected to the control module, is used to receive control commands and drive the upper belt conveyor to adjust its position until the upper belt conveyor and the lower belt conveyor are aligned.

[0017] The beneficial effects of this invention are as follows: The alignment control method and system for belt conveyors provided by this invention, by setting a unique visual detection perspective below the discharge end of the upper belt conveyor, realizes automatic identification and coordinate calibration of the edge of the counterweight block of the lower conveyor, thereby constructing a complete automatic alignment closed-loop control scheme. This method replaces the traditional method that relies on manual observation and adjustment, greatly improving alignment accuracy and efficiency, and fundamentally solving the problem of material spillage caused by misalignment; at the same time, the closed-loop logic ensures that the system can continuously maintain the alignment state and automatically compensate for minor offsets caused by vibration and other reasons during operation, realizing a leap from static manual alignment to dynamic automatic maintenance, and improving the long-term operational stability of the entire conveying system. Attached Figure Description

[0018] Figure 1 A flowchart illustrating the alignment control method for a belt conveyor provided in this embodiment; Figure 2 A schematic diagram of the planar position of the belt conveyor provided for an embodiment; Figure 3 A three-dimensional positional diagram of the belt conveyor provided for an embodiment; Figure 4 This is a schematic diagram illustrating the coordinate relationship of feature points in the image spatial domain, provided in an embodiment. Figure 5 A schematic diagram illustrating the relationship between the coordinates of feature points in the image spatial domain and rotation correction, provided in this embodiment; Figure 6 This is a schematic diagram of the alignment control system of the belt conveyor provided in the embodiment; Explanation of reference numerals in the attached figures: 1-Upper-level belt conveyor; 2-Vision inspection module; 3-Upper-level conveyor belt; 4-Counterweight; 5-Lower-level belt conveyor. Detailed Implementation

[0019] Currently, the series operation of multi-stage belt conveyors heavily relies on manual alignment and periodic inspections. This method not only suffers from significant subjective visual errors and low adjustment efficiency, but also carries the risk of continuous material spillage and equipment blockage due to inaccurate alignment. To achieve automatic alignment of multi-stage belt conveyors and improve alignment accuracy and efficiency, this invention proposes a technical solution.

[0020] This invention transforms the complex spatial alignment problem into a machine vision-based, quantifiable planar coordinate system problem of tracking and registering geometric feature points. Specifically, a fixed observation coordinate system is cleverly established by installing a vision inspection module parallel to the belt below the discharge end of the upper-level belt conveyor. The vision inspection module's viewpoint can simultaneously capture the lower surface of the upper-level belt conveyor, the counterweight of the lower-level belt conveyor, and the ground at the bottom of the lower-level belt conveyor. The counterweight of the lower-level belt conveyor is selected as the ideal feature reference due to its regular, robust, and well-defined geometric structure. Once the equipment is installed, the vision inspection module and the upper-level belt conveyor become a rigid unit. Therefore, any offset of the lower-level belt conveyor relative to the upper-level belt conveyor will be reflected as a movement of the counterweight's feature position in the image acquired by the vision inspection module. During initial installation, fine-tuning is used to achieve the ideal alignment between the two belt conveyors. At this point, an image recognition algorithm extracts and permanently records the image position coordinates of two easily trackable target feature points on the counterweight: the lower left edge vertex B1 and the lower right edge vertex B2. The line connecting these two points, B1B2, defines a baseline, representing the spatial relationship of perfect alignment in the image. During equipment operation, image acquisition continues. For each new acquired image frame, the same image processing procedure as in the calibration phase is executed, re-identifying the same two feature points on the counterweight in a complex background: the real-time lower left edge vertex A1 and the real-time lower right edge vertex A2. Finally, the real-time acquired coordinates of A1 and A2 are compared with the stored baseline coordinates of B1 and B2, the offset is calculated, and corresponding control commands are generated based on the offset to drive the upper-level belt conveyor to perform the corresponding position adjustment. After adjustment, the vision detection module immediately acquires images again, repeats the above steps, calculates the new offset, and performs position adjustments until the upper-level and lower-level belt conveyors are aligned.

[0021] The technical solutions in this embodiment will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0022] Figure 1 A flowchart illustrating an alignment control method for a belt conveyor is shown. Please refer to [link / reference]. Figure 1 The method includes the following steps: Step 1: Install a vision inspection module below the discharge end of the upper belt conveyor. The camera of the vision inspection module is installed parallel to the upper belt conveyor and acquires images that simultaneously include the surface of the upper conveyor belt, the edge contour of the counterweight of the lower belt conveyor, and the ground at the bottom of the lower belt conveyor.

[0023] Please see Figure 2In practical applications, the vision inspection module 2 is first fixed to the frame of the upper belt conveyor 1 using a sturdy mounting bracket, and positioned directly below its discharge end. This location should be chosen in an area with minimal vibration and where it is not easily impacted by materials.

[0024] Then, the camera in the visual inspection module 2 is adjusted so that its optical axis is parallel to the length direction of the upper belt conveyor 1, that is, the long side of the camera's field of view is perpendicular to the belt's running direction, to ensure that a complete lateral field of view can be captured. By adjusting the camera's focal length, it is ensured that the single frame image it acquires can simultaneously cover three key areas: The lower surface of the upper conveyor belt 3: used for subsequent belt surface defect detection.

[0025] The entire edge contour of the counterweight block 4 of the lower belt conveyor 5: used for alignment detection.

[0026] The ground area at the bottom of the lower belt conveyor 5: serves as a background reference and helps improve the stability of image processing.

[0027] In practical applications, industrial cameras with protective housings (dustproof and waterproof) are preferred. They can be equipped with supplementary lights or near-infrared cameras and matching light sources that are not sensitive to visible light can be selected to enhance their ability to resist ambient light interference, depending on the ambient lighting conditions.

[0028] Step 2: Before the equipment is installed and used, adjust the upper belt conveyor to the target alignment position. Based on the image acquired by the vision detection module, identify and obtain the position coordinates of the lower left edge vertex B1 and the lower right edge vertex B2 of the target of the counterweight block of the lower belt conveyor at this time.

[0029] Specifically, this step is performed after the initial system installation or major overhaul, and its purpose is to establish aligned reference coordinates. The implementation method is as follows: First, the operator manually adjusts the upper belt conveyor 1 to a position that is considered to be perfectly aligned with the lower belt conveyor 5, ensuring that the material can be accurately fed into the center of the belt of the lower belt conveyor 5.

[0030] Then, in this state, one or more images are acquired through the visual inspection module 2. Subsequently, based on the image recognition algorithm, the lower left edge vertex of the counterweight block 4 of the lower belt conveyor 5 in the acquired images is accurately identified, and its coordinates are recorded as the lower left edge vertex of the target. Simultaneously, identify its lower right edge vertex and record it as the target's lower right edge vertex. Please see Figure 3 .

[0031] Finally, the position coordinates of B1 and B2 are permanently stored as reference data, and all subsequent alignment judgments will be based on these position coordinates.

[0032] In the above process, the image recognition process based on the image recognition algorithm for the acquired image includes: The acquired images were converted to HSV space and preprocessed using contrast linear transformation. Noise reduction and morphological processing were performed on the images. The Wallner algorithm with adaptive threshold segmentation was used for image binarization segmentation. Connected component contours were extracted, and the edge images of the counterweight blocks were extracted using the Canny edge detection algorithm. Finally, the lower left connection point of the vertical and horizontal lines of the edge images was extracted as B1, and the lower right connection point was extracted as B2.

[0033] Specifically, first, the acquired image (RGB image) is converted to the HSV color space to separate brightness and color information. A contrast linear transformation (adjusting gain and bias) is then used to enhance the difference between the target features and the background. Next, a median filtering algorithm is used to suppress image noise, followed by dilation and erosion operations (morphological opening or closing operations) to eliminate small interference points and connect broken edges. Then, an improved Wallner algorithm based on adaptive thresholding is used for binarization segmentation of the image, better adapting to uneven lighting and dust interference. Next, the Canny edge detection algorithm is used to extract clear edges from the binary image, and then connected component contours matching the size and shape of counterweight block 4 are searched and filtered. Finally, in the found counterweight block contours, its lower left edge vertex is calculated, denoted as B1, and its lower right edge vertex is calculated, denoted as B2.

[0034] The above image recognition process effectively overcomes interference from dust and lighting changes in industrial environments through color space conversion, adaptive noise reduction, and improved edge detection algorithms, achieving stable and accurate extraction of edge feature points of the counterweight.

[0035] Step 3: During equipment operation, based on the images acquired by the vision detection module, the real-time position coordinates of the lower edge vertex of the counterweight block of the lower belt conveyor are identified and obtained in real time. The real-time lower edge vertex includes the real-time lower left edge vertex A1 and the real-time lower right edge vertex A2.

[0036] Specifically, this step is executed cyclically during normal equipment operation to achieve real-time monitoring. The implementation method is as follows: The visual inspection module 2 continuously acquires real-time images at a preset frequency (e.g., 10Hz). For each frame of the acquired image, based on the same image recognition algorithm as in step 2, it accurately identifies the lower left edge vertex of the counterweight block 4 of the lower belt conveyor 5 in the acquired image and records its coordinates as the real-time lower left edge vertex. Simultaneously, its lower right edge vertex is identified and recorded as the real-time lower right edge vertex. Please see Figure 3 .

[0037] In practical applications, during actual operation of the equipment, due to severe shaking, sudden large amount of dust obstruction, sudden change in lighting or extreme offset angle, the image recognition algorithm may be unable to capture both lower edge vertices (A1 and A2) of the counterweight block 4 at a certain moment, and can only successfully identify one of them.

[0038] To ensure recovery and ultimately accurate alignment even in such abnormal situations, this embodiment further includes the following method when only one point, A1 or A2, is identified: Based on the identified misalignment of the X coordinates between A1 or A2 and B1, the position of the upper belt conveyor 1 is initially adjusted. After adjustment, if A1 and A2 appear simultaneously in the acquired image, proceed to step 4; if only one of A1 and A2 is in the field of view, adjust the position of the upper belt conveyor 1 again according to the identified X-coordinate misalignment between A1 or A2 and B2, until A1 and A2 appear simultaneously in the acquired image.

[0039] In practical applications, the first step is to acquire a single real-time point (A1 or A2, denoted as A, with coordinates as follows). Compare the X-coordinates of A and B1 with the X-coordinates of the reference point B1, and calculate the misalignment of the X-coordinates between A and B1. : Then, according to The value of drives the upper belt conveyor 1 to move horizontally. The direction of movement is determined by _____. The positive or negative determination (for example, if) If the value is positive, move to the right; if the value is negative, move to the left. The moving distance can be taken as... Or a preset fixed step size.

[0040] After completing the initial adjustments, immediately return to step 3 to re-acquire the image and identify the feature points. If the real-time lower left edge vertex A1 and the real-time lower right edge vertex A2 can be identified simultaneously in the new acquired image, exit the adjustment process and proceed to the normal step 4. If, in the new acquired image, only the original A or another point can still be identified, but both points cannot be identified simultaneously, then the X-coordinates of a single real-time point A are compared with those of another reference point B2 to calculate the X-coordinate misalignment between A and B2. : Then, according to The value of drives the upper belt conveyor 1 to move horizontally. Similarly, the direction of movement is changed from . The positive or negative determination (for example, if) If the value is positive, move to the right; if the value is negative, move to the left. The moving distance can be taken as... Or a preset fixed step size.

[0041] Repeat the above steps, checking after each adjustment whether A1 and A2 can be identified simultaneously. Once both points are successfully identified, immediately proceed to step 4 for fine-tuning. Through an iterative strategy based on sequential comparisons of a single point and two reference points (B1 and B2), the counterweight 4 of the lower-level belt conveyor 5 can be gradually guided to the central area of ​​the camera's field of view, thereby restoring complete visual features and creating conditions for subsequent high-precision alignment.

[0042] Step 4: Calculate the offset in real time based on the position coordinates of B1, B2, A1 and A2, and control the upper belt conveyor to adjust its position according to the offset until the upper belt conveyor and the lower belt conveyor are aligned.

[0043] Specifically, this step is the core control logic for achieving automatic alignment. The implementation method is as follows: The real-time acquired coordinates A1 and A2 are compared with the stored reference coordinates B1 and B2 to calculate the precise offset. In this embodiment, the offset includes: rotation offset. Horizontal offset and vertical offset .

[0044] Please see Figure 4 Rotation offset The horizontal offset is the angle between line segment A1A2 and the baseline line segment B1B2 in the image plane. With longitudinal offset This represents the average displacement difference between A1 and B1, and between A2 and B2 in the X and Y axes after correcting for rotational offset.

[0045] In this embodiment, step 4 specifically includes steps 41 to 43: Step 41: Calculate the rotational offset, and control the upper belt conveyor to perform rotational offset correction based on the rotational offset until the rotational offset meets the following requirements: ,in, Indicates the rotation offset. This represents the rotation offset threshold.

[0046] Among them, rotation offset The calculation formula is as follows: ; in, Indicates the position coordinates of A1. Indicates the position coordinates of A2. Indicates the position coordinates of B1. This represents the position coordinates of B2.

[0047] The formula calculates the angles between the real-time line segment A1A2 and the baseline line segment B1B2 and the horizontal axis of the image coordinate system, and then calculates the difference between them to obtain the relative angle between them.

[0048] In practical applications, the rotational offset is calculated. Next, the calculated rotational offset will be... With respect to the preset rotation offset threshold (For example, 0.0087 radians, approximately 0.5°) for comparison. If This drives the frame of the upper-level belt conveyor 1 to rotate around its center or a suitable fulcrum. The direction of rotation is to make... The direction of decrease (e.g., if) If the value is positive, rotate counterclockwise. After each rotation adjustment, immediately re-acquire the image, identify points A1 and A2, and calculate the new value. Value. This loop continues until the value is satisfied. Corrective conditions.

[0049] Step 42: Calculate the horizontal offset, and control the upper belt conveyor to perform horizontal offset correction based on the horizontal offset until the horizontal offset meets the following requirements: ,in, Indicates the horizontal offset. This indicates the horizontal offset threshold.

[0050] Please see Figure 5 Since the upstream belt conveyor 1 may have rotated relative to the calibration state, the coordinates of real-time point A1 are... It cannot be directly compared with the reference point B1; it is necessary to calculate the real-time point A1 and perform rotational offset correction. Position coordinates : ; .

[0051] After rotational offset correction The position coordinates, the horizontal offset The calculation formula is as follows: .

[0052] This formula is obtained by taking... The average deviation of B1, A2, and B2 in the X direction improves the stability of the calculation. With horizontal offset threshold (For example, 5 pixel units) are compared. If Then, control the upper belt conveyor 1 to move left or right. The distance (or controlled proportionally) until... .

[0053] Step 43: Calculate the longitudinal offset, and control the upper belt conveyor to perform longitudinal offset correction based on the longitudinal offset until the longitudinal offset meets the following requirements: ,in, Indicates the vertical offset. This indicates the vertical offset threshold.

[0054] After rotational offset correction The position coordinates, the longitudinal offset The calculation formula is as follows: .

[0055] Similarly, this formula improves stability through averaging. With vertical offset threshold (For example, 5 pixel units) are compared. If Then, control the upstream belt conveyor 1 to move forward or backward. The distance, until .

[0056] The above adjustment process first completes the rotational offset correction, then performs horizontal and vertical offset corrections. After each translation adjustment, the system may need to fine-tune the rotation. Therefore, in the specific implementation, a large loop can be set up to sequentially adjust the rotation. , , The alignment is considered complete only when all three parameters simultaneously meet their respective threshold requirements, through a process of detection and adjustment. This step-by-step decoupling control strategy effectively avoids mutual interference between adjustments of multiple degrees of freedom, ensuring the convergence and reliability of the system.

[0057] Furthermore, existing belt conveyors typically only possess basic material conveying functions and lack the ability to monitor the health of the conveyor belt itself in real time. For defects such as surface wear, longitudinal cracks, loose joints, localized tears, and belt misalignment that are prone to occur during long-term, high-load operation, the common approach is to conduct periodic shutdowns for manual inspection. This method not only interrupts continuous operation and affects production efficiency, but manual inspection also carries the risk of missed inspections and misjudgments. If some potential and developing defects are not detected and addressed in a timely manner, they are highly likely to expand during operation, ultimately leading to catastrophic accidents such as belt breakage, causing production interruptions, material contamination, and huge economic losses.

[0058] To address the above problems, the method provided in this embodiment further includes: Images of belt defects, including surface wear, cracks, loose joints, and misalignment, are collected in advance. Machine learning algorithms are then used to train the samples to generate a defect recognition model. The visual inspection module synchronously acquires surface and edge images of the upper conveyor belt; after grayscale conversion and noise reduction preprocessing, the defect recognition model is called to analyze the image and identify whether a preset defect type exists; if a defect is identified, the early warning module is triggered to issue an alarm, and the defect type, location, and severity information are displayed on the display module.

[0059] In practical applications, a large number of known defect sample images of various types are pre-collected, and machine learning algorithms (such as CNN convolutional neural networks) are used to learn the visual feature patterns of different defects (such as wear, cracks, etc.) from these samples, generating a defect recognition model capable of summarizing and identifying these features. During equipment operation, the same vision hardware used for alignment detection synchronously acquires real-time images of the belt surface. After preprocessing these images to improve their quality, they are fed into the pre-trained defect recognition model for analysis. The model matches and compares the real-time images with the learned knowledge base to determine the presence, type, and severity of defects. Once confirmed, an alarm mechanism is immediately activated.

[0060] The above process transforms the accident handling model from reactive maintenance to proactive early warning, enabling timely detection of defects in their nascent or early stages. This effectively prevents catastrophic accidents such as belt breakage caused by the expansion of small defects, ensuring production safety. Furthermore, it achieves continuous automated inspection, completely overcoming the shortcomings of manual inspection in terms of frequency, accuracy, and reliability, freeing up manpower, and significantly improving the automation and intelligence level of equipment management.

[0061] In summary, the alignment control method for belt conveyors provided in this embodiment completely replaces the traditional method relying on manual observation and adjustment by installing a vision inspection module and setting up an automated identification, calculation, and adjustment process. This not only greatly reduces the labor intensity of operators and avoids visual errors caused by subjective human judgment, but also achieves alignment results far exceeding manual precision through precise coordinate calculation and closed-loop control. This fundamentally eliminates material spillage caused by alignment deviations, reducing cleaning costs and the risk of equipment jamming. By pre-calibrating and establishing stable reference coordinates, during operation, corresponding feature points are identified in real time and compared with the reference coordinates for calculation, ultimately driving the actuator to make adjustments. This closed-loop logic ensures continuous alignment and automatically compensates for minor offsets caused by vibration and other factors during operation, achieving a leap from static manual alignment to dynamic automatic maintenance, and improving the long-term operational stability of the entire conveying system. Furthermore, this embodiment also integrates real-time belt defect monitoring on the same hardware platform, laying a solid technical foundation for subsequent predictive maintenance and comprehensively improving equipment safety management.

[0062] Based on the above technical solution, this embodiment also proposes an alignment control system for a belt conveyor, used to implement the alignment control method for the belt conveyor described in the embodiment. Please refer to [link to relevant documentation]. Figure 6 The system includes: The vision inspection module is fixedly installed below the discharge end of the upper belt conveyor. The camera of the vision inspection module is installed parallel to the upper belt conveyor. It is used to collect images that simultaneously include the surface of the upper conveyor belt, the edge contour of the counterweight of the lower belt conveyor, and the ground at the bottom of the lower belt conveyor. The control module, communicatively connected to the vision detection module, is used to adjust the upper-level belt conveyor to the target alignment position before the equipment is installed and used. Based on the image acquired by the vision detection module, it identifies and obtains the position coordinates of the lower left edge vertex B1 and the lower right edge vertex B2 of the counterweight block of the lower-level belt conveyor at this time. During equipment operation, based on the image acquired by the vision detection module, it identifies and obtains the real-time position coordinates of the lower edge vertex of the counterweight block of the lower-level belt conveyor, including the real-time lower left edge vertex A1 and the real-time lower right edge vertex A2. The offset is calculated in real time based on the position coordinates of B1, B2, A1, and A2, and corresponding control commands are generated based on the offset. The drive execution module, connected to the control module, is used to receive control commands and drive the upper belt conveyor to adjust its position until the upper belt conveyor and the lower belt conveyor are aligned.

[0063] It is understood that since the alignment control system of the belt conveyor described in this embodiment is a system for implementing the alignment control method of the belt conveyor described in the embodiment, the system disclosed in the embodiment is relatively simple to describe because it corresponds to the method disclosed in the embodiment. For relevant parts, please refer to the description of the method, and it will not be repeated here.

Claims

1. An alignment control method for a belt conveyor, characterized in that, The method includes: Step 1: Install a vision inspection module below the discharge end of the upper belt conveyor. The camera of the vision inspection module is installed parallel to the upper belt conveyor and acquires images that simultaneously include the surface of the upper conveyor belt, the edge contour of the counterweight of the lower belt conveyor, and the ground at the bottom of the lower belt conveyor. Step 2: Before the equipment is installed and used, adjust the upper belt conveyor to the target alignment position. Based on the image acquired by the vision detection module, identify and obtain the position coordinates of the lower left edge vertex B1 and the lower right edge vertex B2 of the target of the counterweight block of the lower belt conveyor at this time. Step 3: During equipment operation, based on the images acquired by the vision detection module, the real-time position coordinates of the lower edge vertex of the counterweight block of the lower belt conveyor are identified and obtained in real time. The real-time lower edge vertex includes the real-time lower left edge vertex A1 and the real-time lower right edge vertex A2. Step 4: Calculate the offset in real time based on the position coordinates of B1, B2, A1 and A2, and control the upper belt conveyor to adjust its position according to the offset until the upper belt conveyor and the lower belt conveyor are aligned.

2. The alignment control method for a belt conveyor according to claim 1, characterized in that, In step 3, when only one point between A1 and A2 is identified, the method further includes: Based on the identified misalignment of the X coordinates between A1 or A2 and B1, the position of the upper belt conveyor is initially adjusted. After adjustment, if A1 and A2 appear simultaneously in the acquired image, proceed to step 4; if only one of A1 and A2 is in the field of view, adjust the position of the upper belt conveyor again according to the identified X-coordinate misalignment between A1 or A2 and B2, until A1 and A2 appear simultaneously in the acquired image.

3. The alignment control method for a belt conveyor according to claim 1, characterized in that, The offset includes rotational offset, horizontal offset, and vertical offset.

4. The alignment control method for a belt conveyor according to claim 3, characterized in that, Step 4 specifically includes: Calculate the rotational offset, and control the upstream belt conveyor to perform rotational offset correction based on the rotational offset until the rotational offset meets the following requirements: ,in, Indicates the rotation offset. Indicates the rotation offset threshold; Calculate the horizontal offset, and control the upstream belt conveyor to perform horizontal offset correction based on the horizontal offset until the horizontal offset meets the following requirements: ,in, Indicates the horizontal offset. Indicates the horizontal offset threshold; Calculate the longitudinal offset, and control the upstream belt conveyor to perform longitudinal offset correction based on the longitudinal offset until the longitudinal offset meets the following requirements: ,in, Indicates the vertical offset. This indicates the vertical offset threshold.

5. The alignment control method for a belt conveyor according to claim 4, characterized in that, The method for calculating the rotational offset includes: Calculate the angle between the line connecting A1 and A2 and the line connecting B1 and B2; this angle is the rotation offset. The calculation formula is as follows: ； in, Indicates the position coordinates of A1. Indicates the position coordinates of A2. Indicates the position coordinates of B1. This represents the position coordinates of B2.

6. The alignment control method for a belt conveyor according to claim 5, characterized in that, The horizontal offset The calculation formula is as follows: ； in, The X-coordinate of A1 after rotational offset correction is represented by the following formula: 。 7. The alignment control method for a belt conveyor according to claim 5, characterized in that, The vertical offset The calculation formula is as follows: ； in, The Y-coordinate of A1 after rotational offset correction is represented by the following formula: 。 8. The alignment control method for a belt conveyor according to claim 1, characterized in that, The image recognition process includes: The acquired images are converted to HSV space and preprocessed using contrast linear transformation. Noise reduction and morphological processing are performed on the images. The Wallner algorithm with adaptive threshold segmentation is used for image binarization segmentation. Connected component contours are extracted, and the edge images of the counterweight blocks are extracted using the Canny edge detection algorithm. Finally, the lower left connection point of the vertical and horizontal lines of the edge images is extracted as A1 or B1, and the lower right connection point is extracted as A2 or B2.

9. The alignment control method for a belt conveyor according to claim 1, characterized in that, The method further includes: Images of belt defects, including surface wear, cracks, loose joints, and misalignment, are collected in advance. Machine learning algorithms are then used to train the samples to generate a defect recognition model. The visual inspection module synchronously acquires surface and edge images of the upper conveyor belt; after grayscale conversion and noise reduction preprocessing, the defect recognition model is called to analyze the image and identify whether a preset defect type exists; if a defect is identified, the early warning module is triggered to issue an alarm, and the defect type, location, and severity information are displayed on the display module.

10. An alignment control system for a belt conveyor, characterized in that, For implementing the alignment control method for a belt conveyor as described in any one of claims 1 to 9, the system comprises: The vision inspection module is fixedly installed below the discharge end of the upper belt conveyor. The camera of the vision inspection module is installed parallel to the upper belt conveyor. It is used to collect images that simultaneously include the surface of the upper conveyor belt, the edge contour of the counterweight of the lower belt conveyor, and the ground at the bottom of the lower conveyor belt. The control module, communicatively connected to the vision detection module, is used to adjust the upper-level belt conveyor to the target alignment position before the equipment is installed and used. Based on the image acquired by the vision detection module, it identifies and obtains the position coordinates of the lower left edge vertex B1 and the lower right edge vertex B2 of the counterweight block of the lower-level belt conveyor at this time. During equipment operation, based on the image acquired by the vision detection module, it identifies and obtains the real-time position coordinates of the lower edge vertex of the counterweight block of the lower-level belt conveyor, including the real-time lower left edge vertex A1 and the real-time lower right edge vertex A2. The offset is calculated in real time based on the position coordinates of B1, B2, A1, and A2, and corresponding control commands are generated based on the offset. The drive execution module, connected to the control module, is used to receive control commands and drive the upper belt conveyor to adjust its position until the upper belt conveyor and the lower belt conveyor are aligned.

Images