A conveyor belt deviation correction detection method and system based on artificial intelligence

Abstract

The present invention relates to the field of conveyor detection, and more specifically, the present invention relates to a conveyor belt deviation correction detection method and system based on artificial intelligence, the method comprising: obtaining the torque data of the head roller and the tail roller of the conveyor, respectively calculating the torque difference between the head and the tail, and obtaining the torque balance factor. According to the torque balance factor of the conveyor, the minimum value is taken as the overall balance index of the conveyor. The change trend of the torque balance factor is analyzed, and the mutation index at each moment is obtained, so as to identify the existing anomalies. At the same time, the historical images of the conveyor are processed, marked and segmented, the edge sequence is extracted, and the Euclidean distance between the edges is calculated to evaluate the degree of deformation, and a deviation recognition network is constructed. The mutation index and the use time are input, and the deviation state of the conveyor is output, so as to realize comprehensive monitoring of the operating status of the conveyor. The present invention can quickly respond to abnormal conditions of the conveyor by calculating the mutation index in real time.

Description

Technical Field

[0001] The present invention relates to the field of conveyor detection, and more specifically, to a conveyor belt deviation correction detection method and system based on artificial intelligence. Background Art

[0002] Conveyor deviation correction is a key link in maintaining the normal operation of the conveying system. A deviated conveyor will cause excessive edge wear, damage to the equipment structure, inaccurate material delivery, material overflow and safety accidents, thus affecting production efficiency and operational safety. Regular deviation correction operations can ensure that materials are stably delivered to the designated location, reduce material losses, avoid environmental pollution, and reduce equipment failure rates and maintenance costs.

[0003] The existing Chinese patent application document with publication number CN116788783A discloses a belt conveyor deviation control method and system based on fuzzy control algorithm, including: monitoring the working status of the belt, continuously transmitting the value to the PLC control module and the frequency converter; obtaining the deviation value according to the fuzzy rule base, and outputting the control signal to the frequency converter; the frequency converter controls the running speed and angle of the deviation adjustment device motor to correct the belt to return to the normal position. The invention communicates with other belt conveyor deviation correction subsystems through the data transmission substation, expands into a multi-belt intelligent deviation correction system, and is centrally controlled by the intelligent management platform. The PLC uses the fuzzy control algorithm to infinitely close-loop control the correction roller and the correction roller's correction wheel servo actuator to make angle adjustment based on the detected belt conveyor deviation, and through reasonable dead zone control, the belt conveyor deviation is controlled within the set value range. And it will not cause frequent swing adjustment due to slight system deviations or changes.

[0004] In the above scheme, the deviation correction detection method is combined with fuzzy control algorithm and PLC control technology. However, the fuzzy control system may have a response lag, which may cause the deviation correction action to be not timely enough, resulting in hidden dangers in the conveyor transmission, thereby affecting production efficiency. Summary of the invention

[0005] In order to solve the problem that the deviation correction detection has a response lag, which may cause the deviation correction action to be not timely enough and lead to hidden dangers in the conveyor transportation, the present invention provides solutions in the following aspects.

[0006] In the first aspect, a conveyor belt deviation correction detection method based on artificial intelligence includes: obtaining torque data of the head roller and the tail roller of the conveyor, obtaining the difference of the torque data on both sides of the head roller of the conveyor, so as to determine the torque balance factor on both sides of the head roller, and similarly obtaining the torque balance factor on both sides of the tail roller; respectively taking the smallest torque balance factor of the head and tail torque balance factors as the torque balance factor of the conveyor, obtaining the continuous change coefficient of the torque balance factor of the conveyor at each moment, and determining whether the torque balance factor has a continuous downward change according to the amplitude of the continuous change coefficient, so as to obtain the mutation index at each moment; obtaining a historical conveyor image, marking the conveyor image, and performing image segmentation on the marked conveyor image, obtaining a marked area, extracting and arranging the long edge of each marked area, obtaining an edge sequence, calculating the Euclidean distance of adjacent long edge positions, obtaining a distance sequence, and obtaining the deformation degree of the conveyor; constructing a conveyor deviation recognition network, inputting the mutation index and the conveyor usage time obtained in real time at the current moment into the deviation recognition network, and outputting the result of identifying the abnormality of the conveyor belt.

[0007] The effect is as follows: by acquiring the torque data of the head and tail rollers of the conveyor in real time, the torque balance factor is calculated using the torque data, and the torque balance state of the head and tail rollers of the conveyor is accurately evaluated. By analyzing the continuous change coefficient of the torque balance factor, it is possible to identify whether the torque balance factor has a continuous downward change, thereby timely discovering potential abnormal trends; combining the marking and image segmentation technology of historical conveyor images, extracting edge sequences and calculating the Euclidean distance, effectively evaluating the deformation degree of the conveyor, and the constructed conveyor deviation recognition network can combine real-time data with historical data, and intelligently identify whether the conveyor has deviated, thereby improving the accuracy of prediction and the timeliness of response.

[0008] Preferably, the torque balance factor on both sides of the head roller includes:

[0009] The maximum and minimum absolute values ​​of the difference of the torque data at both sides of the head roller at all times are obtained, the ratio of the minimum value to the maximum value is calculated, and the torque balance factor on both sides of the head roller is obtained.

[0010] The effect is: by obtaining the absolute value of the torque data difference on both sides of the head roller at all times, the difference in the torque of the rollers on both sides can be quantified, thereby evaluating the balance state between them, and calculating the ratio of the minimum value to the maximum value as the torque balance factor, which provides an indicator to measure the torque balance on both sides of the head roller. Timely detection of torque imbalance can prevent the conveyor from deviating due to uneven tension on both sides, reducing the risk of equipment damage and production interruption.

[0011] Preferably, the step of obtaining the continuous variation coefficient of the torque balance factor of the transmission machine at each moment includes:

[0012] The difference between the torque balance factors at two consecutive moments is calculated, and the change direction of the torque balance factor is determined according to the difference. In response to the difference being greater than zero, the continuous change coefficient of the torque balance factor at the current moment is 1, and in response to the difference being less than or equal to zero, the continuous change coefficient of the current torque balance factor is 0.

[0013] The effect is that by calculating the difference between the torque balance factors at two consecutive moments, the changing trend of the torque balance factor can be analyzed to determine whether the operating status of the conveyor is continuously improving or deteriorating; determining the direction of change of the torque balance factor helps to identify whether the conveyor is tending towards a more balanced or more unbalanced state.

[0014] Preferably, obtaining the mutation index at each moment includes:

[0015] Get the Time to The sequence of torque balance factors at the moment, calculate the standard deviation of the sequence of torque balance factors, and obtain the Time to At each moment, calculate The product of the continuous variation coefficients of the standard deviation is multiplied by the negative exponent of the standard deviation and the product of the continuous variation coefficients. The mutation index at the moment.

[0016] Preferably, the deformation degree of the conveyor satisfies the following relationship:

[0017] ;

[0018] In the formula, Indicates the degree of deformation, Indicates The first half of the edge sequence, Indicates The second half of the edge sequence, Indicates The distance sequence between the edge of a strip and the adjacent previous edge, Indicates The distance sequence between the edge of a strip and the adjacent previous edge, represents the maximum value function, represents the mean value function, Represents the standard deviation function.

[0019] Preferably, the deviation identification network is a bp fully connected neural network, wherein the training set of the preset neural network is a two-dimensional vector consisting of the transmission machine usage time in the historical data, the mutation index and the corresponding transmission machine usage time, wherein the transmission machine usage time is the total usage time up to the moment of the mutation coefficient, and the mutation index that deviates from the training set is marked, and the training set carrying the mark is put into the fully connected neural network. In response to the loss function being less than the loss threshold or the number of training times reaching the preset number of training times, the training is completed to obtain the transmission machine deviation identification network.

[0020] Preferably, the loss function of the deviation recognition network satisfies the following relationship:

[0021] ;

[0022] In the formula, represents the loss function of the deviation recognition network, Indicates The moment corresponds to the degree of deformation of the image. The label value representing the marked area, represents the logarithmic function, Indicates The probability that the moment prediction is off.

[0023] In the second aspect, a conveyor belt deflection detection system based on artificial intelligence includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned conveyor belt deflection detection method based on artificial intelligence is implemented.

[0024] The present invention has the following effects:

[0025] 1. The present invention calculates the standard deviation of the torque balance factor to obtain the degree of fluctuation of the torque balance factor within a period of time, and uses the size of the mutation index as an early warning signal to indicate the possible risk of conveyor deviation. By calculating the mutation index in real time, the abnormal situation of the conveyor can be responded to quickly, thereby improving the response speed and processing efficiency of the entire system.

[0026] 2. The present invention can further improve the accuracy of deviation correction detection by taking the degree of deformation as the weight of the loss function in the deviation recognition network, so that the model pays more attention to samples with larger deformation degrees, thereby improving the accuracy of prediction. Through intelligent recognition and early warning, maintenance plans can be optimized, emergency repairs caused by deviation can be reduced, and maintenance costs can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0028] Figure 1 It is a method flow chart of steps S1 to S4 in a conveyor belt deviation correction detection method based on artificial intelligence in an embodiment of the present invention.

[0029] Figure 2 It is a structural block diagram of a conveyor belt deviation correction detection system based on artificial intelligence in an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0031] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0032] Reference Figure 1 , a conveyor belt deviation correction detection method based on artificial intelligence includes steps S1 to S4, which are as follows:

[0033] S1: Obtain the torque data of the head roller and the tail roller of the conveyor, and obtain the difference of the torque data on both sides of the head roller of the conveyor to determine the torque balance factor on both sides of the head roller. Similarly, obtain the torque balance factor on both sides of the tail roller.

[0034] It is further explained that torque sensors are installed on the rollers at the head and tail of the conveyor to monitor and record the torque data of the rollers on both sides in real time, and the torque data of the head and tail rollers are collected synchronously.

[0035] It should be noted that as the use time increases, the tension of the conveyor itself will decrease. At this time, when heavy objects come into contact with the conveyor, the tension of the conveyor may be inconsistent on both sides, or the tension in different sections may be uneven. The conveyor may deviate to the side with less tension, causing the torque of the rollers on the left and right sides of the conveyor to change significantly. Therefore, it is necessary to calculate the torque balance factor on both sides of the rollers at the head and tail of the conveyor. The specific steps are as follows:

[0036] The maximum and minimum absolute values ​​of the difference of the torque data at both sides of the head roller at all times are obtained, the ratio of the minimum value to the maximum value is calculated, and the torque balance factor on both sides of the head roller is obtained.

[0037] Specifically, the torque balance factor of the head roller satisfies the following relationship:

[0038] ;

[0039] In the formula, Indicates The torque balance factor of the moment conveyor head, Indicates The torque data of the left roller of the conveyor head at all times, Indicates The torque data of the roller on the right side of the transmission head is transmitted at all times.

[0040] By torque balance factor The closer the ratio is to 1, the more balanced the torque of the rollers on both sides is; the farther away from 1, the higher the degree of imbalance is.

[0041] The difference between the torque balance factors at two consecutive moments is calculated, and the change direction of the torque balance factor is determined according to the difference. In response to the difference being greater than zero, the continuous change coefficient of the torque balance factor at the current moment is 1; in response to the difference being less than or equal to zero, the continuous change coefficient of the current torque balance factor is 0.

[0042] Specifically, the continuous variation coefficient satisfies the following relationship:

[0043] ;

[0044] In the formula, Indicates The continuous change coefficient at any time, Indicates The torque balance factor of the moment conveyor head, Indicates The torque balance factor of the momentary transmission head.

[0045] That is, the torque balance factor at the current moment is Less than the previous moment , then the continuous variation coefficient Assigned a value of 1, indicating that the torque balance factor is The torque balance factor at the current moment is continuously reduced. Greater than or equal to the previous moment , then the continuous variation coefficient Assigned a value of 0, indicating that the torque balance factor is The time increases continuously.

[0046] S2: The smallest torque balance factor of the head and tail torque balance factors is used as the torque balance factor of the transmission machine, and the continuous change coefficient of the torque balance factor of the transmission machine at each moment is obtained. According to the amplitude of the continuous change coefficient, it is determined whether the torque balance factor has a continuous downward change, so as to obtain the mutation index at each moment.

[0047] Get the Time to The sequence of torque balance factors at the moment, calculate the standard deviation of the sequence of torque balance factors, and obtain the Time to At each moment, calculate The product of the continuous variation coefficients of the standard deviation is multiplied by the negative exponent of the standard deviation and the product of the continuous variation coefficients. The mutation index at the moment.

[0048] Specifically, the mutation coefficient satisfies the following relationship:

[0049] ;

[0050] In the formula, Indicates Moment mutation index, represents the standard deviation function, Indicates The torque balance factor of the moment conveyor, Indicates The torque balance factor of the moment conveyor, Indicates The continuous changing coefficient at each moment.

[0051] In this embodiment, setting The default value is 5, which means that the analysis The torque balance factor sequence and the corresponding continuous variation coefficient of the five moments before the moment, where the standard deviation reflects the Time to The fluctuation degree of the internal torque balance factor at the moment, the larger the standard deviation, the more drastic the fluctuation of the torque balance factor; Reflects the torque balance factor in the Time to The changes within a moment. Time continuous change coefficient If it is continuously 1, it means that the torque balance factor continues to decrease, then the product term will increase, otherwise, the product term will decrease.

[0052] S3: Obtain historical conveyor images, mark the conveyor images, and perform image segmentation on the marked conveyor images to obtain the marked areas, extract the long edges of each marked area and arrange them to obtain an edge sequence, calculate the Euclidean distance between adjacent long edge positions, obtain a distance sequence, and obtain the deformation degree of the conveyor.

[0053] In this embodiment, an industrial camera is arranged just above the head and tail of the conveyor to obtain the conveyor video, extract frames, obtain the conveyor image, and use a semantic segmentation network to perform image segmentation. This technology is well known to those skilled in the art and will not be described in detail. Before image segmentation, the pixels of the conveyor area in the historical conveyor image are coded and marked as 1, and the pixels of the area other than the conveyor area are coded and marked as 0. The semantic segmentation network (CNN) is put into training. In response to the loss function value being less than 0.001 or reaching the preset number of training times, the training is completed to obtain a label recognition network. The image is input into the label recognition network to obtain a label area image (0-1 image), and a label area is obtained.

[0054] Specifically, the deformation degree of the conveyor satisfies the following relationship:

[0055] ;

[0056] In the formula, Indicates the degree of deformation, Indicates The first half of the edge sequence, Indicates The second half of the edge sequence, Indicates The distance sequence between the edge of a strip and the adjacent previous edge, Indicates The distance sequence between the edge of a strip and the adjacent previous edge, represents the maximum value function, represents the mean value function, Represents the standard deviation function.

[0057] In other words, by analyzing the position changes of the conveyor edge sequence and the consistency of the distance between the edges, the maximum deformation degree of the conveyor is comprehensively evaluated.

[0058] S4: Construct a conveyor deviation identification network, input the mutation index and conveyor usage time obtained in real time at the current moment into the deviation identification network, and output the abnormal result of identifying the conveyor tape.

[0059] The deviation identification network is a bp fully connected neural network, wherein the training set of the preset neural network is a two-dimensional vector consisting of the transmission machine usage time in the historical data, the mutation index and the corresponding transmission machine usage time, wherein the transmission machine usage time is the total usage time up to the moment of the mutation coefficient, and the mutation index that deviates from the training set is marked, and the training set carrying the mark is put into the fully connected neural network. In response to the loss function being less than the loss threshold or the number of training times reaching the preset number of training times, the training is completed to obtain the transmission machine deviation identification network.

[0060] It should be noted that the BP (Backpropagation) fully connected neural network is used to train the model, and the deformation degree is introduced as a weight into the loss function, which can enhance the deviation recognition network's ability to recognize severe deformation.

[0061] Specifically, the loss function of the deviation recognition network satisfies the following relationship:

[0062] ;

[0063] In the formula, represents the loss function of the deviation recognition network, Indicates The moment corresponds to the degree of deformation of the image. The label value representing the marked area, represents the logarithmic function, Indicates The probability that the moment prediction is off.

[0064] The present invention also provides a conveyor belt deviation correction detection system based on artificial intelligence. Figure 2 As shown, the system includes a processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an artificial intelligence-based conveyor belt deviation correction detection method according to the first aspect of the present invention is implemented.

[0065] The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

[0066] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0067] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0068] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A conveyor belt deviation correction detection method based on artificial intelligence, characterized in that: include: Obtain torque data of the head roller and the tail roller of the conveyor, and obtain the difference of the torque data on both sides of the head roller of the conveyor to determine the torque balance factor on both sides of the head roller, and similarly obtain the torque balance factor on both sides of the tail roller; The smallest torque balance factor of the torque balance factors of the head and the tail is respectively used as the torque balance factor of the transmission machine, and the continuous change coefficient of the torque balance factor of the transmission machine at each moment is obtained; according to the amplitude of the continuous change coefficient, it is determined whether the torque balance factor has a continuous downward change, so as to obtain the mutation index at each moment, including: obtaining the first Time to The sequence of torque balance factors at the moment, calculate the standard deviation of the sequence of torque balance factors, and obtain the Time to At each moment, calculate The product of the continuous variation coefficients of the standard deviation is multiplied by the negative exponent of the standard deviation and the product of the continuous variation coefficients. The mutation index at the moment; Obtain a historical conveyor image, mark the conveyor image, perform image segmentation on the marked conveyor image, obtain the marked area, extract and arrange the long edge of each marked area to obtain an edge sequence, calculate the Euclidean distance between adjacent long edge positions to obtain a distance sequence, and obtain the deformation degree of the conveyor; Construct a conveyor deviation identification network, input the mutation index and conveyor usage time obtained in real time at the current moment into the deviation identification network, and output the result of identifying conveyor tape abnormalities.

2. According to the artificial intelligence-based conveyor belt deviation correction detection method of claim 1, it is characterized in that: The torque balance factors on both sides of the head roller include: The maximum and minimum absolute values ​​of the difference of the torque data at both sides of the head roller at all times are obtained, the ratio of the minimum value to the maximum value is calculated, and the torque balance factor on both sides of the head roller is obtained.

3. The artificial intelligence-based conveyor belt deviation correction detection method according to claim 1 is characterized in that: The step of obtaining the continuous variation coefficient of the torque balance factor of the transmission machine at each moment includes: The difference between the torque balance factors at two consecutive moments is calculated, and the change direction of the torque balance factor is determined according to the difference. In response to the difference being greater than zero, the continuous change coefficient of the torque balance factor at the current moment is 1, and in response to the difference being less than or equal to zero, the continuous change coefficient of the current torque balance factor is 0.

4. The artificial intelligence-based conveyor belt deviation correction detection method according to claim 1 is characterized in that: The deformation degree of the conveyor satisfies the following relationship: ； In the formula, Indicates the degree of deformation, Indicates The first half of the edge sequence, Indicates The second half of the edge sequence, Indicates The distance sequence between the edge of a strip and the adjacent previous edge, Indicates The distance sequence between the edge of a strip and the adjacent previous edge, represents the maximum value function, represents the mean value function, Represents the standard deviation function.

5. The artificial intelligence-based conveyor belt deviation correction detection method according to claim 1 is characterized in that: The deviation identification network is a bp fully connected neural network, wherein the training set of the preset neural network is a two-dimensional vector consisting of the transmission machine usage time in the historical data, the mutation index and the corresponding transmission machine usage time, wherein the transmission machine usage time is the total usage time up to the moment of the mutation coefficient, and the mutation index that deviates from the training set is marked, and the training set carrying the mark is put into the fully connected neural network. In response to the loss function being less than the loss threshold or the number of training times reaching the preset number of training times, the training is completed to obtain the transmission machine deviation identification network.

6. The artificial intelligence-based conveyor belt deviation correction detection method according to claim 1 is characterized in that: The loss function of the deviation recognition network satisfies the following relationship: ； In the formula, represents the loss function of the deviation recognition network, Indicates The moment corresponds to the degree of deformation of the image. The label value representing the marked area, represents the logarithmic function, Indicates The probability that the moment prediction is off.

7. A conveyor belt deviation correction detection system based on artificial intelligence, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the conveyor belt deviation correction detection method based on artificial intelligence according to any one of claims 1 to 6 is implemented.

Images