Belt conveyor closed-loop protection system and method

Abstract

The present application proposes a belt conveyor closed-loop protection system and method, which relates to the field of belt conveyor operation protection technology, and the method includes: real-time acquisition of video images on the belt conveyor; based on the video images, combining deep learning methods and statistical theories to calculate coal particle size, coal pile conditions, and coal quantity data; through a deviation prediction model, the coal particle size, coal pile conditions, and coal quantity data are calculated to predict the degree of deviation of the belt conveyor; according to the coal particle size, coal pile conditions, coal quantity data, and deviation degree, the parameters of the correction device, crushing device, or plug plate are adjusted or the belt conveyor is locked. The present application makes up for the problem that the prior art only identifies but does not control, fundamentally solves the protection problems of belt conveyor deviation and coal pile, increases the service life of the belt conveyor, and reduces the workload of personnel in handling problems.

Description

Technical Field

[0001] The present application relates to the technical field of belt conveyor operation protection, and in particular to a belt conveyor closed-loop protection system and method. Background Art

[0002] Belt conveyors are widely used in coal mining, cement manufacturing, fertilizer production, and grain processing as key equipment for efficiently handling bulk materials. Although originally designed to achieve smooth and continuous material transfer, practical challenges remain. The most prominent issues include belt deviation (i.e., "deviation") and material accumulation (such as coal). These problems can easily cause material leakage or dispersion during transportation, resulting in resource waste; they can also accelerate aging and damage to the conveyor belt surface material, posing a threat to the overall performance of the equipment. Therefore, it is crucial to take effective measures to prevent and resolve these issues to ensure safe and stable production line operation and extend the service life of mechanical equipment.

[0003] To address belt conveyor deviation and coal accumulation issues and ensure safe and stable operation, various methods have been proposed, including contact sensors and non-contact visual inspection. Contact-based inspections often employ vertical roller-type deviation detection, which offers the advantages of simplicity and convenience. However, the mechanical structure rubs against the belt conveyor, making it susceptible to damage. Non-contact inspections often utilize lasers and video. However, existing solutions primarily rely on various means to detect belt conveyor deviation, but none address the root causes of deviation or intelligently address the consequences. Summary of the Invention

[0004] The present application aims to solve one of the technical problems in the related art at least to a certain extent.

[0005] To this end, the first purpose of this application is to propose a closed-loop protection method for a belt conveyor.

[0006] The second objective of this application is to provide a belt conveyor closed-loop protection system.

[0007] The third objective of this application is to provide an electronic device.

[0008] The fourth object of this application is to provide a computer-readable storage medium.

[0009] A fifth object of this application is to provide a computer program product.

[0010] To achieve the above-mentioned objectives, the first embodiment of the present application proposes a closed-loop protection method for a belt conveyor, comprising:

[0011] Acquire video images on belt conveyor in real time;

[0012] Based on the video images, coal particle size, coal pile conditions, and coal quantity data are calculated by combining deep learning methods and statistical theory;

[0013] The deviation prediction model calculates the coal particle size, the coal pile condition, and the coal quantity data to predict the deviation degree of the belt conveyor. The deviation prediction model uses a support vector machine model as a basic framework, adopts a radial basis function as the kernel function of the support vector machine model, and uses an ant colony optimization algorithm to optimize the penalty parameters and kernel function parameters of the support vector machine model;

[0014] According to the coal particle size, the coal pile condition, the coal quantity data, and the degree of deviation, the parameters of the deviation correction device, the crushing device, or the inserting plate are adjusted or the belt conveyor is locked.

[0015] Optionally, adjusting the parameters of the correction device, the crushing device, or the inserting plate or locking the belt conveyor according to the coal particle size, the coal pile condition, the coal quantity data, and the degree of deviation includes:

[0016] When the deviation degree is greater than the preset deviation degree threshold, the parameters of the correction device are adjusted;

[0017] When the coal particle size is greater than a preset coal particle size threshold, the parameters of the crushing device are adjusted;

[0018] When the coal quantity data is greater than the preset coal quantity threshold, the parameters of the plug-in board are adjusted to reduce the coal quantity;

[0019] When the coal pile level is greater than a preset coal pile level threshold, the belt conveyor is locked.

[0020] Optionally, the process of calculating the coal particle size based on the video image includes:

[0021] Initializing a matching relationship between the size of the actual coal block and the pixel size in the video image, determining the actual physical size represented by each pixel in the video image, determining a conversion ratio between pixels and real-world units, and determining a relationship between the actual size of the coal block and the pixel size based on the conversion ratio;

[0022] using the coal particle size recognition model trained and saved in the offline stage to identify the coal lumps in the video frame of the video image, and calculating the number and particle size of the coal lumps;

[0023] Calculate the coal particle size on the belt conveyor based on the calculated number and particle size of the coal pieces , the calculation formula is:

[0024]

[0025] in, are the number and weight of coal particle size greater than 200 pixels, are the number and weight of coal particle size detected between 100 and 200 pixels, are the amount and weight of the number of hours when the coal particle size is 80~100 pixels, are the number and weight of coal particle size detection when it is 50~80 pixels, are the number and weight of coal particle size less than 50 pixels when detected;

[0026] The process of calculating the coal quantity data based on the video image includes:

[0027] The video image is sequentially preprocessed, a region of interest is acquired, coal flow features are extracted, and an image segmentation process is performed using a video-based multi-feature coal quantity detection model;

[0028] Compare the area of ​​the coal flow region obtained after processing with the area of ​​the region of interest, and use the comparison results as coal quantity data , the expression is:

[0029]

[0030] in, is the area of ​​the identified coal flow region, is the area of ​​the region of interest;

[0031] The process of calculating the coal pile situation based on the video image includes:

[0032] Based on the production process relationship of the belt conveyor, a coal quantity and coal stacking model is established. For the scenario where two belt conveyors converge into one belt conveyor, the actual coal quantity correlation relationship is defined as:

[0033]

[0034] in, They are respectively the coal quantity of the belt conveyors detected in real time by the two belt conveyors before and after the convergence. is based on The predicted amount of coal from the belt conveyor;

[0035] According to the amount of coal conveyor Calculate the predicted belt conveyor coal quantity after converging , and calculate the degree of deviation, and use the degree of deviation as the coal pile situation , the expression is:

[0036]

[0037] in, They are The historical maximum and minimum values.

[0038] Optionally, the offline training phase of the coal particle size recognition model includes:

[0039] Collecting a preset number of historical image data sets of the belt conveyor during operation, and dividing the historical image data sets into a training set and a test set according to a preset ratio, wherein the historical image data sets include images under various interference conditions;

[0040] Data annotation is performed on images with different interference conditions, and model training parameters of the YOLOv8 network model are set;

[0041] The YOLOv8 network model is trained based on the annotated historical image dataset. After the training is completed, the coal particle size recognition model is obtained.

[0042] Optionally, the deviation prediction model uses a support vector machine model as a basic framework, and the offline training process of the deviation prediction model includes:

[0043] Obtain a sample data set, and divide the sample data set into a training set and a test set according to a preset ratio, and , sample coal quantity data , Sample coal pile situation , sample coal particle size increment , sample coal quantity data increment , sample coal pile situation increase As the input of the support vector machine model, the actual deviation degree as the predicted label;

[0044] Performing preprocessing and normalization on the sample data set, wherein the preprocessing includes data cleaning and data standardization;

[0045] The radial basis function is used as the kernel function of the support vector machine model, and the penalty parameter of the support vector machine model is optimized by the ant colony optimization algorithm. and kernel function parameters Optimize and combine cross-validation technology to simulate the processed sample training set data as the control variable to evaluate the performance of the support vector machine model under different parameter combinations and set the optimal penalty parameter and kernel function parameters The support vector machine model below is used as the deviation prediction model.

[0046] Optionally, the actual deviation degree The acquisition process is:

[0047] The rollers are divided into left and right rollers according to the center point of the belt conveyor area, and the deviation direction is calculated according to the average area of ​​the left and right rollers; the deviation degree of the rollers is calculated based on the deviation direction and the deviation judgment threshold.

[0048] The edge is divided into left and right edges based on the center point of the belt conveyor area. A fixed value is set. The degree of edge deviation is calculated based on the corresponding values ​​of the left and right edge lines when not deviating, the corresponding values ​​of the left and right edge lines, and the deviation judgment threshold.

[0049] The average value of the roller deviation degree and the edge deviation degree is taken as the actual deviation degree.

[0050] Optionally, the radial basis function is used as the kernel function of the support vector machine model, and the penalty parameter of the support vector machine model is optimized by the ant colony optimization algorithm. and kernel function parameters Optimize and combine cross-validation technology to simulate the processed sample training set data as the control variable to evaluate the performance of the support vector machine model under different parameter combinations and set the optimal penalty parameter and kernel function parameters The following support vector machine model is used as the deviation prediction model, including:

[0051] Initialize the parameters of the ant colony optimization algorithm, including setting the number of ant colonies, the maximum number of iterations, the pheromone evaporation coefficient and the pheromone volatility factor, and setting the penalty parameter and kernel function parameters Search scope;

[0052] Search for the best penalty parameter within a given range using the ant colony optimization algorithm parameter algorithm and kernel function parameters , the ant colony moves according to the current pheromone concentration and continuously updates the pheromone concentration on the path to gradually approach the optimal solution;

[0053] In each iteration, the support vector machine model is trained using the current parameter combination, and the mean square error and mean absolute percentage error of the processed sample training set data are calculated and used as fitness evaluation indicators;

[0054] Check whether the number of iterations of the support vector machine model reaches the maximum number of iterations or whether the fitness value meets the preset accuracy requirements. If not, continue to perform parameter search; otherwise, stop searching and output the optimal penalty parameter found. and kernel function parameters .

[0055] Optionally, also include:

[0056] The goal of supporting vector machine model training is to obtain a regression model so that the predicted value and the true value The error between them is the smallest. The relevant expressions of the support vector machine model include minimizing the objective function, constraints, Lagrangian function and regression model, as follows:

[0057] Minimize:

[0058]

[0059] Constraints:

[0060]

[0061] Lagrangian function:

[0062]

[0063] Regression Model:

[0064]

[0065] in, is the weight vector, is the slack variable, For the The slack variables corresponding to the training samples are is a constant, is the kernel function, For the The feature vector of the training samples, For the The true value of the training samples, For the The predicted value of the training samples, is the Lagrange multiplier, is an additional Lagrange multiplier, For the Additional Lagrange multipliers corresponding to training samples are used to handle equality constraints , is the bias term, the superscript represents a conjugate variable, It is The Lagrange multiplier corresponding to the training samples, is the number of training samples.

[0066] Optionally, the predicted value of the deviation prediction model is subjected to denormalization processing to obtain the predicted deviation degree of the belt conveyor.

[0067] To achieve the above-mentioned purpose, the second embodiment of the present application proposes a belt conveyor closed-loop protection system, comprising:

[0068] Video acquisition module, used to acquire video images on the belt conveyor in real time;

[0069] A video analysis module is used to calculate coal particle size, coal pile status, and coal quantity data based on the video images in combination with deep learning methods and statistical theory;

[0070] a deviation prediction module, configured to calculate the coal particle size, the coal pile condition, and the coal quantity data through a deviation prediction model to predict the degree of deviation of the belt conveyor. The deviation prediction model uses a support vector machine model as a basic framework, adopts a radial basis function as the kernel function of the support vector machine model, and uses an ant colony optimization algorithm to optimize the penalty parameters and kernel function parameters of the support vector machine model;

[0071] A control execution module, configured to adjust parameters of a deviation correction device, a crushing device, or a plug-in plate, or lock the belt conveyor according to the coal particle size, the coal pile condition, the coal quantity data, and the deviation degree;

[0072] Optionally, adjusting the parameters of the correction device, the crushing device, or the inserting plate or locking the belt conveyor according to the coal particle size, the coal pile condition, the coal quantity data, and the degree of deviation includes:

[0073] When the deviation degree is greater than the preset deviation degree threshold, the parameters of the correction device are adjusted;

[0074] When the coal particle size is greater than a preset coal particle size threshold, the parameters of the crushing device are adjusted;

[0075] When the coal quantity data is greater than the preset coal quantity threshold, the parameters of the plug-in board are adjusted to reduce the coal quantity;

[0076] When the coal pile level is greater than a preset coal pile level threshold, the belt conveyor is locked.

[0077] To achieve the above-mentioned purpose, a third embodiment of the present application provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor;

[0078] The memory stores computer-executable instructions;

[0079] The processor executes the computer-executable instructions stored in the memory to implement the method as described in any one of the first aspects.

[0080] To achieve the above-mentioned purpose, the fourth embodiment of the present application proposes a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by a processor, they are used to implement the method as described in any one of the first aspects.

[0081] To achieve the above-mentioned objectives, the fifth embodiment of the present application proposes a computer program product, which implements any one of the methods in the first aspect when executed by a processor.

[0082] The belt conveyor closed-loop protection system and method, electronic equipment and storage medium provided in this application focus on analyzing the factors that cause belt conveyor accidents, establishing a belt conveyor prediction model under various factors to predict the deviation situation, and performing on-site correction through a correction device. After the correction, the belt conveyor correction feedback is identified in real time in combination with video, thereby forming a closed-loop protection process, which makes up for the problem that the existing technology only recognizes but does not control, fundamentally solves the protection problems of belt conveyor deviation and coal piling, increases the service life of the belt conveyor, and reduces the workload of personnel in handling problems.

[0083] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0085] Figure 1 A schematic flow chart of a closed-loop protection method for a belt conveyor provided in an embodiment of the present application;

[0086] Figure 2 A schematic flow chart of the coal particle size identification process provided in an embodiment of the present application;

[0087] Figure 3 A schematic diagram of a scenario in which two belt conveyors converge into one belt conveyor during coal pile condition detection provided in an embodiment of the present application;

[0088] Figure 4 A schematic diagram of a process flow for performing predictions using the deviation prediction model provided in an embodiment of the present application;

[0089] Figure 5 This is a structural schematic diagram of a belt conveyor closed-loop protection system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0090] The following describes in detail embodiments of the present application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0091] Belt conveyors are widely used in coal mining, cement manufacturing, fertilizer production, and grain processing as key equipment for efficiently handling bulk materials. Although originally designed to achieve smooth and continuous material transfer, practical challenges remain. The most prominent issues include belt deviation (i.e., "deviation") and material accumulation (such as coal). These problems can easily cause material leakage or dispersion during transportation, resulting in resource waste; they can also accelerate aging and damage to the conveyor belt surface material, posing a threat to the overall performance of the equipment. Therefore, it is crucial to take effective measures to prevent and resolve these issues to ensure safe and stable production line operation and extend the service life of mechanical equipment.

[0092] To address belt conveyor deviation and coal accumulation issues and ensure safe and stable operation, various methods have been proposed, including contact sensors and non-contact visual inspection. Contact-based inspections often employ vertical roller-type deviation detection, which offers the advantages of simplicity and convenience. However, these mechanical structures rub against the belt conveyor, making them susceptible to damage. Non-contact inspections often utilize lasers and video. However, existing solutions typically rely on various means to detect belt conveyor deviation, but none address the root causes of deviation or intelligently address the consequences.

[0093] Belt conveyors currently suffer from common problems such as coal accumulation and deviation. Current video and control systems are unable to address the root causes of these problems and establish a closed-loop protection system. Furthermore, the primary cause of deviation is excessive or insufficient material flow, which results in uneven tension on the conveyor and leads to deviation. Material accumulation on the conveyor can lead to an uneven surface, causing deviation. Uneven material particle size can cause material to shift during conveyance, leading to belt deviation.

[0094] To address this issue, the present invention provides a closed-loop protection method for a belt conveyor. Figure 1 This is a flow chart of a closed-loop protection method for a belt conveyor provided in an embodiment of the present application. Figure 1 As shown, the method includes the following steps:

[0095] Step 101: Acquire video images on the belt conveyor in real time.

[0096] In an embodiment of the present application, an industrial camera can be used to capture high-definition video images of the belt conveyor.

[0097] Step 102: Calculate the coal particle size, coal pile condition, and coal quantity data based on the video image in combination with deep learning methods and statistical theory.

[0098] Figure 2 The flow chart of the coal particle size identification process provided in the embodiment of the present application can be used as a reference.

[0099] In the embodiment of the present application, the process of calculating the coal particle size based on the video image includes the following steps:

[0100] Step 201: Initialize the matching relationship between the size of the actual coal block and the pixel size in the video image.

[0101] In this embodiment, the actual physical size of each pixel in the image must first be determined. This is typically accomplished through a camera calibration process, which involves the use of a calibration plate to determine the conversion ratio between image pixels and real-world units (e.g., meters or centimeters). The range of possible sizes of coal lumps in the image is then determined, and the relationship between the actual size of the coal lumps and the pixel size is determined based on the conversion ratio.

[0102] Step 202: Use the coal particle size recognition model trained and saved in the offline stage to identify the coal lumps in the video frame of the video image, and calculate the number and particle size of the coal lumps.

[0103] In an embodiment of the present application, the offline training stage of the coal particle size recognition model includes: collecting a preset number of historical picture data sets of the belt conveyor during operation, dividing the historical picture data sets into a training set and a test set according to a preset ratio, wherein, when making the data set, it is necessary to consider the completeness of the data, that is, the historical picture data set needs to include pictures under various interference conditions; data annotation is performed on the pictures under different interference conditions, and the model training parameters of the YOLOv8 network model are set; the YOLOv8 network model is trained according to the annotated historical picture data set, and after the training is completed, a coal particle size recognition model is obtained.

[0104] In a possible embodiment, 31,200 historical images are collected from 201 belt conveyors as a training set, and 4,680 images are collected as a test set. The training and test sets are not repeated, and the interference conditions include daytime, nighttime, dust, fog and other interferences.

[0105] In addition, the configured model training parameters include epochs, batch-size, mini-batch, data, and img-size. The parameter configuration is related to the hardware configuration, such as epochs=300, batch-size=64, and img-size=640×640.

[0106] The meaning of each parameter is:

[0107] epochs: In machine learning, an epoch is when the entire training dataset is fully trained on the model. Setting the epochs parameter determines how many times the model will train on the training dataset.

[0108] Batch-size: This refers to the number of samples fed into the model for training at each iteration. A smaller batch-size provides more detailed gradient updates but may slow down training. A larger batch-size speeds up training but may lead to less stable training.

[0109] Mini-batch: This is another way of saying batch-size, which refers to the small group of examples used for training in each iteration. In practice, mini-batch gradient descent is a commonly used optimization algorithm that uses a small batch of examples each time it updates the model parameters.

[0110] Data: In the context of model training, data usually refers to the dataset used to train the model. This may include input features and corresponding labels or target values.

[0111] img-size: In image processing and computer vision tasks, img-size refers to the size of the input image, usually in pixels. When training a coal particle size recognition model, you need to specify the width and height of the image so that the model can correctly process the input data.

[0112] It should be noted that this application saves the coal particle size recognition model after offline training as an online application model, and this application tests the coal particle size recognition model on a test data set, and its recognition accuracy is 99.2%.

[0113] Step 203: Calculate the coal particle size on the belt conveyor based on the calculated number and particle size of the coal blocks. , the calculation formula is:

[0114]

[0115] in, are the number and weight of coal particle size greater than 200 pixels, are the number and weight of coal particle size detected between 100 and 200 pixels, are the amount and weight of the number of hours when the coal particle size is 80~100 pixels, are the number and weight of coal particle size detection when it is 50~80 pixels, are the number and weight of coal particle size less than 50 pixels.

[0116] It can be understood that the coal particle size detection interval can also be set and changed according to the actual scenario, and the value given in the above coal particle size detection interval is not used as a limitation of the coal particle size detection interval of this application.

[0117] It is understandable that different pixel weights need to be set according to the actual scenario, and this application does not make any specific limitations on this.

[0118] In one possible embodiment, , , , , .

[0119] In the embodiment of the present application, the process of calculating the coal quantity data based on the video image is implemented by a multi-feature coal quantity detection model based on the video, and the process specifically includes the following steps:

[0120] Step 301: Preprocess the video image, obtain the region of interest, extract the coal flow features, and perform image segmentation processing in sequence through a video-based multi-feature coal quantity detection model.

[0121] In the embodiment of the present application, the preprocessing process may include denoising, enhancement, resizing and normalization processes. Since the preprocessing process is a conventional technical means in this field, the present application will not describe and explain it in detail. Region of interest acquisition refers to determining the position of the coal flow in the image through image analysis technology (such as edge detection, color segmentation), and cropping the area containing the coal flow from the original image to reduce irrelevant information and improve processing efficiency. Coal flow feature extraction refers to: extracting the texture features of the coal flow area; extracting the shape features of the coal blocks, such as area, perimeter, aspect ratio, etc.; extracting the color features of the coal blocks, such as statistical values ​​in RGB and HSV color spaces; if the video sequence contains time information, the motion features of the coal blocks, such as speed, direction, etc., can be extracted. Image segmentation refers to distinguishing the coal blocks from the background, and methods such as threshold segmentation and cluster segmentation can be used.

[0122] Step 302: Compare the area of ​​the coal flow region obtained after processing with the area of ​​the region of interest, and use the comparison result as the coal quantity data. , the expression is:

[0123]

[0124] in, is the area of ​​the identified coal flow region, is the area of ​​the region of interest.

[0125] Figure 3 The schematic diagram of the scenario in which two belt conveyors converge into one belt conveyor during coal pile condition detection provided in the embodiment of the present application can be used as a reference.

[0126] In the embodiment of the present application, the process of calculating the coal pile situation based on the video image includes:

[0127] Step 401: Based on the production process relationship of the belt conveyor, a coal quantity and coal pile model is established, such as Figure 3 As shown in the figure, for the scenario where two belt conveyors converge into one belt conveyor, the coal quantity association relationship in actual situation is defined as:

[0128]

[0129] in, They are respectively the coal quantity of the belt conveyors detected in real time by the two belt conveyors before and after the convergence. is based on Predicted amount of coal for the belt conveyor.

[0130] It should be noted that for the scenario where two belt conveyors converge into one belt conveyor, ideally, the relationship is based on the coal quantity; However, due to the camera installation angle and location, this is not true.

[0131] Step 402: According to the amount of coal from the belt conveyor Calculate the predicted belt conveyor coal quantity after converging , and calculate the degree of deviation, and use the degree of deviation as the coal pile situation , the expression is:

[0132]

[0133] in, They are The historical maximum and minimum values.

[0134] In the embodiment of the present application, in real-time application, the coal quantity of the belt conveyor is calculated in real time. , and the real-time calculated data Enter the formula , thus calculating amount.

[0135] Step 103 , using a deviation prediction model to calculate the coal particle size, coal pile condition, and coal quantity data, and predict the deviation degree of the belt conveyor.

[0136] In the embodiment of the present application, deviation detection is a novel and efficient belt deviation detection method and system. By combining roller fixed points with image processing technology, it realizes accurate monitoring of the belt deviation status. This model is basically consistent with the coal particle size detection model and is divided into offline training and online detection processes. During the offline training process, pictures are collected, edges and rollers are marked, and during the data labeling process, data of different belt conveyors and deviation degrees in different environments are mainly considered. During the online detection stage, the belt conveyor video is collected in real time, initialized by matching the actual roller size with the pixel size, and the deviation model saved by offline training is called for real-time detection. The position and number of rollers and the position and number of belt conveyor edges are obtained in real time.

[0137] Figure 4 The flowchart of the deviation prediction model provided in the embodiment of the present application can be used as a reference. In the embodiment of the present application, the deviation prediction model is to establish a correlation model between coal particle size, coal quantity, coal pile and belt conveyor deviation. The deviation prediction model uses the support vector machine model as the basic framework, adopts the radial basis function as the kernel function of the support vector machine model, and uses the ant colony optimization algorithm to optimize the penalty parameters and kernel function parameters of the support vector machine model. Specifically, the offline training process of the deviation prediction model includes:

[0138] Step 501: Obtain a sample data set, and divide the sample data set into a training set and a test set according to a preset ratio. , sample coal quantity data , Sample coal pile situation , sample coal particle size increment , sample coal quantity data increment , sample coal pile situation increase As the input of the support vector machine model, the actual deviation degree as the predicted label.

[0139] In the embodiment of the present application, the coal particle size, coal quantity, coal pile and deviation value calculation method have been described above. , sample coal quantity data , Sample coal pile situation The particle size of the sample coal in steps 101-102 can be , sample coal quantity data , Sample coal pile situation The sample coal particle size increment is obtained by the acquisition process , sample coal quantity data increment , sample coal pile situation increase The particle size of the sample coal before and after a period of time is , sample coal quantity data , Sample coal pile situation To calculate the actual deviation degree It is obtained by the degree of roller deviation and edge deviation.

[0140] In one possible embodiment, 420 sets of data are randomly shuffled and divided into 336 training sets and 84 test sets in a ratio of 4:1. , , thereby ensuring the randomness and authenticity of the data.

[0141] Specifically, in the embodiment of the present application, based on the position and number of the rollers and the position and number of the edge of the belt conveyor, the actual degree of deviation is obtained. The acquisition process includes:

[0142] Step 601: Divide the idlers into left and right idlers according to the center point of the belt conveyor area, calculate the deviation direction according to the average area of ​​the left and right idlers, and calculate the deviation degree of the idlers based on the deviation direction and the deviation judgment threshold.

[0143] In the embodiment of the present application, the roller position is divided into left roller ztg and right roller ytg according to the center point of the belt conveyor area, and then the area of ​​each roller is calculated. , and calculate the average area of ​​the left and right rollers , and use this to calculate the deviation direction, the formula is as follows:

[0144]

[0145] As a possible implementation method, historical data is used to calculate the deviation judgment threshold , the expression is:

[0146]

[0147] Where, is the number of rollers, and the subscript indicates the number of the rollers, for example, Refers to the area of ​​the first left idler, Refers to the area of ​​the first right idler.

[0148] In the embodiment of the present application, the process of calculating the area of ​​each roller and the average area of ​​the left and right rollers is a conventional technical solution in the field, which needs to be determined according to the geometric shape of the rollers in the actual scene. For example, the area of ​​the left and right rollers is calculated using the circular area calculation formula, and the average area is calculated based on the area of ​​all left rollers and the number of left rollers and the area of ​​all right rollers and the number of right rollers. This application will not elaborate on this. Therefore, the calculation expression for the degree of roller deviation is:

[0149]

[0150] Where, The degree of roller deviation.

[0151] Step 602: Divide the edge into a left edge and a right edge according to the center point of the belt conveyor area, set a fixed value, and calculate the degree of edge deviation based on the corresponding values ​​of the left edge line and the right edge line when they are not deviated, the corresponding values ​​of the left edge line and the right edge line, and the deviation judgment threshold.

[0152] In the embodiment of the present application, similar to step 601, the edge position is divided into the left edge zby and the right edge yby according to the center point of the tape machine area, and a fixed value is set. , and calculate a fixed The corresponding value when the belt conveyor does not deviate , and then use the left edge center point to fit the left and right edge lines 、 , and by a fixed value Calculate the left edge line and the right edge corresponding to 、 The above calculation processes are all conventional technical contents in this field.

[0153] In the embodiment of the present application, the fixed value Indicates a certain position of the tape conveyor in the longitudinal direction (usually the length direction). For example, Y=100 may mean that the state of the tape is observed at a position 100 meters away from the starting point of the tape. By fixing this Y value, you can focus on analyzing whether the lateral position of the tape at this longitudinal position meets the expectations. If the tape conveyor is not deviating, the left and right edges of the tape should maintain a specific lateral position at this fixed Y value, and the relative positions of the left and right edges should remain symmetrical and the distance from the center point should be consistent. If the tape is not deviating, then and They should be symmetrical with the center point of the belt conveyor, and the distance between them should meet the design requirements of the belt width.

[0154] In addition, the left and right edge lines are fitted using the left edge center point 、 , and by a fixed value Calculate the left edge line and the right edge corresponding to 、 This is also the conventional technical content in this field.

[0155] For example, the coordinates of the center point of the left edge can be used to fit the left edge line and the right edge line through curve fitting technology. 、 Common fitting methods include polynomial fitting, spline fitting, etc., which are not specifically limited in this application.

[0156] Then, since the equations of the left and right edge lines are obtained by fitting, when a fixed value is given (such as a fixed Y coordinate value), the fixed value can be substituted into the equations of the left and right edge lines, and the corresponding values ​​can be obtained by reverse solution. (x value of the left edge) and (x value of the right edge)

[0157] Furthermore, the deviation judgment threshold is calculated within the normal range as:

[0158]

[0159] The subscripts represent the edge numbers, for example, refers to the value of the first left edge, Refers to the value of the first right edge.

[0160] Therefore, the calculation expression of the edge deviation degree is:

[0161]

[0162] Where, The degree of edge deviation.

[0163] Step 603: Taking the average of the roller deviation degree and the edge deviation degree as the actual deviation degree.

[0164] In the embodiment of the present application, by integrating the degree of deviation of the roller and edge deviation , take the average value as the final actual deviation degree .

[0165] Step 502: Preprocess and normalize the sample data set. Preprocessing includes data cleaning and data standardization.

[0166] It is understandable that preprocessing and normalization are technical means well known to those skilled in the art, and this application will not provide further explanation or description thereof.

[0167] Step 503: Use radial basis function as the kernel function of the support vector machine model and use ant colony optimization algorithm to adjust the penalty parameter of the support vector machine model. and kernel function parameters Optimize and combine cross-validation technology to simulate the processed sample training set data as the control variable to evaluate the performance of the support vector machine model under different parameter combinations and set the optimal penalty parameter and kernel function parameters The support vector machine model below is used as the deviation prediction model.

[0168] In the embodiment of the present application, the parameters of the ant colony optimization algorithm are first initialized, including setting the number of ant colonies, the maximum number of iterations, the pheromone evaporation coefficient and the pheromone volatility factor, and setting the penalty parameter and kernel function parameters The search range is then searched for the best penalty parameter within the given range through the ant colony optimization algorithm parameter algorithm and kernel function parameters The ant colony moves according to the current pheromone concentration and continuously updates the pheromone concentration on the path to gradually approach the optimal solution; and in each iteration, the support vector machine model is trained using the current parameter combination, and the mean square error and mean absolute percentage error on the processed sample training set data are calculated, which are used as fitness evaluation indicators; finally, check whether the number of iterations of the support vector machine model reaches the maximum number of iterations or whether the fitness value meets the preset accuracy requirements. If not, continue to perform parameter search; otherwise, stop searching and output the optimal penalty parameter found and kernel function parameters .

[0169] In one embodiment, the number of ant colonies N = 20, the maximum number of iterations K = 200, the pheromone evaporation coefficient ρ = 0.85, and the pheromone volatility factor =0.8, At the same time, determine the penalty parameter and kernel function parameters The search range is, for example, set to [0.01, 100], which is not specifically limited in this application.

[0170] Furthermore, it is understandable that the maximum number of iterations and the preset accuracy requirements should be set according to the actual scenario and data, and this application does not make any specific limitations on this.

[0171] It should be noted that, in the embodiment of the present application, the goal of the support vector machine model training is to obtain a regression model so that the predicted value and the true value The error between them is the smallest. The relevant expressions of the support vector machine model include minimizing the objective function, constraints, Lagrangian function and regression model, as follows:

[0172] Minimize:

[0173]

[0174] Constraints:

[0175]

[0176] The dual problem Lagrangian function used to build the support vector machine model:

[0177]

[0178] Regression Model:

[0179]

[0180] in, is the weight vector, is the slack variable, For the The slack variables corresponding to the training samples are is a constant, is the kernel function, For the The feature vector of the training samples, For the The true value of the training samples, For the The predicted value of the training samples, is the Lagrange multiplier, is an additional Lagrange multiplier, For the Additional Lagrange multipliers corresponding to training samples are used to handle equality constraints , is the bias term, the superscript represents a conjugate variable, It is The Lagrange multiplier corresponding to the training samples, is the number of training samples.

[0181] In addition, since the data is normalized before prediction, it is also necessary to perform denormalization on the predicted value of the deviation prediction model to obtain the predicted deviation degree of the belt conveyor.

[0182] Step 104 , according to the coal particle size, coal pile condition, coal quantity data, and deviation degree, the parameters of the deviation correction device, crushing device, or inserting plate are adjusted, or the belt conveyor is locked.

[0183] In an embodiment of the present application, when the degree of deviation is greater than a preset deviation threshold, the parameters of the correction device are adjusted; when the coal particle size is greater than a preset coal particle size threshold, the parameters of the crushing device are adjusted; when the coal quantity data is greater than a preset coal quantity threshold, the parameters of the insert plate are adjusted to reduce the coal quantity; when the coal stacking degree is greater than the preset coal stacking degree threshold, the belt conveyor is locked.

[0184] It is understood that the position or angle of the deviation-correcting device can be adjusted through the PLC control system until the detected deviation degree returns to a safe range. For example, the PLC receives sensor signals from coal particle size, coal stacking, coal quantity data, and deviation degree, and adjusts the parameters of the deviation-correcting device, crushing device, or insert plate according to preset thresholds and control logic. For example, when the value detected by the deviation degree sensor exceeds the preset deviation degree threshold, the PLC can trigger the adjustment of the deviation-correcting device; when the value detected by the coal particle size sensor exceeds the preset coal particle size threshold, the PLC can adjust the parameters of the crushing device; when the coal quantity data exceeds the preset coal quantity threshold, the PLC can adjust the parameters of the insert plate to reduce the coal quantity; when the coal stacking degree exceeds the preset coal stacking degree threshold, the PLC can execute the locking operation of the belt conveyor.

[0185] In certain specific cases, if automatic correction alone cannot solve the problem, the system will trigger an alarm to prompt manual intervention.

[0186] As can be understood, this application proposes a closed-loop belt conveyor protection method designed to improve the safety and efficiency of traditional belt conveyors by integrating advanced sensing technology, intelligent control algorithms, and automated adjustment mechanisms. This method not only identifies potential problems such as deviation or coal accumulation but also takes immediate corrective action, effectively extending the life of the equipment and reducing the workload of maintenance personnel.

[0187] To demonstrate the beneficial effects of this closed-loop belt conveyor protection method, the system was tested on two identical belt conveyors, A (control group) and B (experimental group), at a coal preparation plant. Conveyor B employed the closed-loop belt conveyor protection method proposed in this invention and operated continuously for six months. Data on conveyor deviations, coal accumulation, and coal particle size showed that control group A experienced 25 significant deviations over the six-month period. In the experimental group, thanks to its real-time monitoring and automatic correction mechanism, only three minor deviations occurred during the same period, each of which was quickly corrected without further impact. The introduction of the closed-loop belt conveyor protection system not only significantly reduced the incidence of accidents but also significantly lowered operating costs and improved efficiency. This demonstrates the high practical value and development potential of this technical solution, ensuring production safety while also generating significant economic benefits for the enterprise.

[0188] In order to implement the above embodiment, the present application also proposes a belt conveyor closed-loop protection system. Figure 5 Schematic diagram of the structure of a belt conveyor closed-loop protection system 10 provided in an embodiment of the present application. Figure 5 As shown, the system includes:

[0189] The video acquisition module 100 is used to acquire video images on the belt conveyor in real time;

[0190] Video analysis module 200, for calculating coal particle size, coal pile status, and coal quantity data based on video images in combination with deep learning methods and statistical theory;

[0191] The deviation prediction module 300 is used to calculate the coal particle size, coal pile condition, and coal quantity data through a deviation prediction model to predict the degree of deviation of the belt conveyor. The deviation prediction model uses a support vector machine model as the basic framework, adopts a radial basis function as the kernel function of the support vector machine model, and uses an ant colony optimization algorithm to optimize the penalty parameters and kernel function parameters of the support vector machine model;

[0192] The control execution module 400 is used to adjust the parameters of the correction device, the crushing device or the inserting plate or lock the belt conveyor according to the coal particle size, the coal pile condition, the coal quantity data and the degree of deviation.

[0193] Optionally, when the degree of deviation is greater than the preset deviation threshold, adjust the parameters of the correction device; when the coal particle size is greater than the preset coal particle size threshold, adjust the parameters of the crushing device; when the coal quantity data is greater than the preset coal quantity threshold, adjust the parameters of the insert plate to reduce the coal quantity; when the coal stacking degree is greater than the preset coal stacking degree threshold, lock the belt conveyor.

[0194] Suppose a coal mine site has been installed with the belt conveyor closed-loop protection system proposed in this invention, which includes all the components mentioned above. One day, during normal production, the conveyor belt slightly deviates due to uneven coal particles. At this point, the system operates as follows:

[0195] The video acquisition module captures this change and immediately transmits it to the video analysis module;

[0196] The video analysis module quickly identified an abnormal increase in coal particle size and a slight accumulation on the left side of the conveyor belt;

[0197] This information is fed into the pre-trained deviation prediction model in the deviation prediction module, which calculates that the deviation is likely to intensify in the next few minutes.

[0198] After receiving the prediction result, the control execution module immediately issues a command to start the deviation correction device on the left side of the conveyor belt and simultaneously reduces the opening of the feed port insert to reduce the coal flow rate;

[0199] After several fine-tuning steps, the conveyor belt returned to its normal trajectory, avoiding a more serious accident.

[0200] Through such a closed-loop feedback mechanism, not only can problems be discovered in a timely manner, but effective measures can also be taken at the first moment to prevent the situation from worsening, greatly improving the safety and stability of equipment operation.

[0201] In order to implement the above embodiments, the present application also proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method provided by the above embodiments.

[0202] In order to implement the above embodiments, the present application also proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the methods provided by the above embodiments.

[0203] In order to implement the above embodiments, the present application also proposes a computer program product, including a computer program, which implements the methods provided by the above embodiments when executed by a processor.

[0204] The collection, storage, use, processing, transmission, provision and disclosure of user personal information involved in this application are in compliance with relevant laws and regulations and do not violate public order and good morals.

[0205] It is important to note that personal information collected from users should be used for legitimate and reasonable purposes and should not be shared or sold beyond these legitimate uses. Furthermore, such collection / sharing should be conducted only after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign an agreement / authorization that includes the relevant user information before using the feature. Furthermore, any necessary steps must be taken to safeguard and secure access to such personal information and ensure that others with access to personal information comply with its privacy policy and procedures.

[0206] This application contemplates providing implementations that allow users to selectively block the use or access of personal information data. Specifically, this disclosure contemplates providing hardware and / or software to prevent or block access to such personal information data. Risks can be minimized by limiting data collection and deleting data once it is no longer needed. Furthermore, where applicable, such personal information can be de-identified to protect user privacy.

[0207] In the descriptions of the foregoing embodiments, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and features of different embodiments or examples, unless they are mutually inconsistent.

[0208] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. Throughout the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0209] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0210] The logic and / or steps represented in a flowchart or otherwise described herein, for example, can be considered a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" is any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (not exhaustive) of computer-readable media include: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0211] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having logic gate circuits for implementing logical functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0212] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0213] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0214] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A belt conveyor closed-loop protection method, characterized in that: include: Acquire video images on belt conveyor in real time; Based on the video images, coal particle size, coal pile conditions, and coal quantity data are calculated by combining deep learning methods and statistical theories; The coal particle size, the coal pile condition and the coal quantity data are calculated by a deviation prediction model to predict the deviation degree of the belt conveyor. The deviation prediction model uses a support vector machine model as a basic framework, adopts a radial basis function as a kernel function of the support vector machine model, and uses an ant colony optimization algorithm to optimize the penalty parameters and kernel function parameters of the support vector machine model; According to the coal particle size, the coal pile condition, the coal quantity data, and the deviation degree, the parameters of the deviation correction device, the crushing device, or the plug plate are adjusted, or the belt conveyor is locked; The method of adjusting the parameters of the deviation correction device, the crushing device or the plug plate or locking the belt conveyor according to the coal particle size, the coal pile condition, the coal quantity data and the deviation degree includes: When the deviation degree is greater than the preset deviation degree threshold, the parameters of the deviation correction device are adjusted; When the coal particle size is greater than a preset coal particle size threshold, adjusting the parameters of the crushing device; When the coal quantity data is greater than the preset coal quantity threshold, adjust the parameters of the plug board to reduce the coal quantity; When the coal pile level exceeds the preset coal pile level threshold, the belt conveyor is locked. The process of calculating the coal particle size based on the video image includes: Initializing the matching relationship between the size of the actual coal block and the pixel size in the video image, determining the actual physical size represented by each pixel in the video image, determining the conversion ratio between the pixel and the real-world unit, and determining the relationship between the actual size of the coal block and the pixel size according to the conversion ratio; Using the coal particle size recognition model trained and saved in the offline stage to identify the coal lumps in the video frame of the video image, and calculate the number and particle size of the coal lumps; Calculate the coal particle size on the belt conveyor based on the calculated number and particle size of the coal pieces , the calculation formula is: in, are the number and weight of coal particle size greater than 200 pixels, are the number and weight of coal particle size detection when it is 100~200 pixels, are the number and weight of coal particle size detection when it is 80~100 pixels, They are the number and weight of coal particle size detection when it is 50~80 pixels. They are the number and weight of coal particle size detected when it is less than 50 pixels; The process of calculating the coal quantity data based on the video image includes: The video image is sequentially preprocessed, the region of interest is acquired, the coal flow feature is extracted, and the image segmentation process is performed using a multi-feature coal quantity detection model based on video; Compare the area of ​​the coal flow region obtained after processing with the area of ​​interest, and use the comparison result as coal quantity data , the expression is: in, is the area of ​​the identified coal flow region, is the area of ​​the region of interest; The process of calculating the coal pile situation based on the video image includes: According to the production process relationship of the belt conveyor, a coal quantity and coal stacking model is established. For the scenario where two belt conveyors converge into one belt conveyor, the coal quantity association relationship in the actual situation is defined as: in, They are the coal quantity of the belt conveyors detected in real time by the two belt conveyors before and after the convergence. is based on The predicted amount of coal for the belt conveyor; According to the amount of coal from the belt conveyor Calculate the coal quantity of the belt conveyor predicted by the belt conveyor after convergence , and calculate the degree of deviation, and use the degree of deviation as the coal pile situation , the expression is: in, They are The historical maximum and minimum values.

2. The method according to claim 1, characterized in that: The offline training phase of the coal particle size recognition model includes: Collecting a preset number of historical picture data sets of the belt conveyor during operation, and dividing the historical picture data sets into a training set and a test set according to a preset ratio, wherein the historical picture data sets contain pictures under various interference conditions; Annotate the images with different interference conditions and set the model training parameters of the YOLOv8 network model; The YOLOv8 network model is trained according to the annotated historical image data set, and after the training is completed, the coal particle size recognition model is obtained.

3. The method according to claim 2, characterized in that The deviation prediction model uses a support vector machine model as a basic framework, and the offline training process of the deviation prediction model includes: Obtain a sample data set, and divide the sample data set into a training set and a test set according to a preset ratio, and separate the sample coal particle size in the sample data set into a training set and a test set. , Sample coal quantity data , Sample coal pile situation , Sample coal particle size increment , Sample coal quantity data increment , Sample coal pile situation increment As the input of the support vector machine model, the actual deviation degree as the predicted label; Preprocessing and normalizing the sample data set, wherein the preprocessing includes data cleaning and data standardization; The radial basis function is used as the kernel function of the support vector machine model, and the penalty parameter of the support vector machine model is optimized using the ant colony optimization algorithm. and kernel function parameters Optimize and combine cross-validation technology to simulate the processed sample training set data as the control variable to evaluate the performance of the support vector machine model under different parameter combinations and set the optimal penalty parameter and kernel function parameters The support vector machine model below is used as the deviation prediction model.

4. The method according to claim 3, characterized in that The actual deviation degree The acquisition process is: The rollers are divided into left rollers and right rollers according to the center point of the belt conveyor area, and the deviation direction is calculated according to the average area of ​​the left and right rollers; the deviation degree of the rollers is calculated based on the deviation direction and the deviation judgment threshold; The edge is divided into a left edge and a right edge according to the center point of the belt conveyor area, a fixed value is set, and the degree of edge deviation is calculated according to the corresponding values ​​of the left edge line and the right edge line when not deviated, the corresponding values ​​of the left edge line and the right edge line, and the deviation judgment threshold; The average value of the roller deviation degree and the edge deviation degree is taken as the actual deviation degree.

5. The method according to claim 4, characterized in that The radial basis function is used as the kernel function of the support vector machine model, and the penalty parameter of the support vector machine model is optimized by using the ant colony optimization algorithm. and kernel function parameters Optimize and combine cross-validation technology to simulate the processed sample training set data as the control variable to evaluate the performance of the support vector machine model under different parameter combinations and set the optimal penalty parameter and kernel function parameters The support vector machine model below is used as the deviation prediction model, including: Initialize the parameters of the ant colony optimization algorithm, including setting the number of ant colonies, the maximum number of iterations, the pheromone evaporation coefficient and the pheromone volatilization factor, and setting the penalty parameter and kernel function parameters Search scope; Search for the best penalty parameter within a given range through the ant colony optimization algorithm parameter algorithm and kernel function parameters , the ant colony moves according to the current pheromone concentration and continuously updates the pheromone concentration on the path to gradually approach the optimal solution; In each iteration, the support vector machine model is trained using the current parameter combination, and the mean square error and mean absolute percentage error on the processed sample training set data are calculated as fitness evaluation indicators; Check whether the number of iterations of the support vector machine model reaches the maximum number of iterations or whether the fitness value meets the preset accuracy requirements. If not, continue to perform parameter search; otherwise, stop searching and output the optimal penalty parameter found. and kernel function parameters .

6. The method according to claim 5, characterized in that Also includes: The goal of supporting vector machine model training is to obtain a regression model so that the predicted value and the true value The error between them is the smallest, and the related expressions of the support vector machine model include minimizing the objective function, constraints, Lagrangian function and regression model, as follows: Minimize: Constraints: Lagrangian function: Regression Model: in, is the weight vector, is the slack variable, For the The slack variables corresponding to the training samples are is a constant, is the kernel function, For the The feature vector of the training samples, For the The true value of the training samples, For the The predicted value of the training samples, is the Lagrange multiplier, are additional Lagrange multipliers, For the Additional Lagrange multipliers corresponding to training samples are used to handle equality constraints , is the bias term, the superscript represents a conjugate variable, It is The Lagrange multiplier corresponding to the training samples is: is the number of training samples.

7. The method according to claim 1, characterized in that The predicted value of the deviation prediction model is subjected to denormalization processing to obtain the predicted deviation degree of the belt conveyor.

8. A belt conveyor closed-loop protection system, characterized in that: include: Video acquisition module, used to acquire video images on the belt conveyor in real time; A video analysis module is used to calculate the coal particle size, coal pile condition and coal quantity data based on the video image in combination with a deep learning method and statistical theory; A deviation prediction module is used to calculate the coal particle size, the coal pile condition, and the coal quantity data through a deviation prediction model to predict the deviation degree of the belt conveyor. The deviation prediction model uses a support vector machine model as a basic framework, adopts a radial basis function as a kernel function of the support vector machine model, and uses an ant colony optimization algorithm to optimize the penalty parameters and kernel function parameters of the support vector machine model; A control execution module, used for adjusting the parameters of the deviation correction device, the crushing device or the plug plate or locking the belt conveyor according to the coal particle size, the coal pile condition, the coal quantity data and the deviation degree; The method of adjusting the parameters of the deviation correction device, the crushing device or the plug plate or locking the belt conveyor according to the coal particle size, the coal pile condition, the coal quantity data and the deviation degree includes: When the deviation degree is greater than the preset deviation degree threshold, the parameters of the deviation correction device are adjusted; When the coal particle size is greater than a preset coal particle size threshold, adjusting the parameters of the crushing device; When the coal quantity data is greater than the preset coal quantity threshold, adjust the parameters of the plug board to reduce the coal quantity; When the coal pile level is greater than a preset coal pile level threshold, the belt conveyor is locked; The process of calculating the coal particle size based on the video image includes: Initializing the matching relationship between the size of the actual coal block and the pixel size in the video image, determining the actual physical size represented by each pixel in the video image, determining the conversion ratio between the pixel and the real-world unit, and determining the relationship between the actual size of the coal block and the pixel size according to the conversion ratio; Using the coal particle size recognition model trained and saved in the offline stage to identify the coal lumps in the video frame of the video image, and calculate the number and particle size of the coal lumps; Calculate the coal particle size on the belt conveyor based on the calculated number and particle size of the coal pieces , the calculation formula is: in, are the number and weight of coal particle size greater than 200 pixels, are the number and weight of coal particle size detection when it is 100~200 pixels, are the number and weight of coal particle size detection when it is 80~100 pixels, They are the number and weight of coal particle size detection when it is 50~80 pixels. They are the number and weight of coal particle size detected when it is less than 50 pixels; The process of calculating the coal quantity data based on the video image includes: The video image is sequentially preprocessed, the region of interest is acquired, the coal flow feature is extracted, and the image segmentation process is performed using a multi-feature coal quantity detection model based on video; Compare the area of ​​the coal flow region obtained after processing with the area of ​​interest, and use the comparison result as coal quantity data , the expression is: in, is the area of ​​the identified coal flow region, is the area of ​​the region of interest; The process of calculating the coal pile situation based on the video image includes: According to the production process relationship of the belt conveyor, a coal quantity and coal stacking model is established. For the scenario where two belt conveyors converge into one belt conveyor, the coal quantity association relationship in the actual situation is defined as: in, They are the coal quantity of the belt conveyors detected in real time by the two belt conveyors before and after the convergence. is based on The predicted amount of coal for the belt conveyor; According to the amount of coal from the belt conveyor Calculate the coal quantity of the belt conveyor predicted by the belt conveyor after convergence , and calculate the degree of deviation, and use the degree of deviation as the coal pile situation , the expression is: in, They are The historical maximum and minimum values.

Images