Fault detection method

Abstract

The invention provides a fault detection method, and relates to the technical field of fault detection, and the method comprises the steps: obtaining a belt image of conveying equipment, and recognizing each damaged region in the belt image; segmenting and extracting each damaged area from the belt image to obtain a plurality of segmented images; determining the damage degree and the fault type of the conveying equipment based on the damage characteristics of the segmented image; wherein the damage characteristics comprise the damage area and / or the damage number. The damage characteristic of the belt is obtained through the belt image, the damage degree and the fault type of the conveying equipment are finally determined, the influence of human factors is avoided, the accuracy of judging the fault type of the conveying equipment is improved, and the situation of misjudgment or missed judgment is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of fault identification, and more particularly, to a fault detection method. Background Art

[0002] As a common mechanical device for transporting materials, belt conveyors are widely used in industries such as mines, metallurgy, building materials, and chemicals. They are characterized by high efficiency, stability, and large-scale transportation. However, it is inevitable that belt conveyors will be damaged as the service time increases and the working environment becomes more complex. The damage situation will affect the continuity and safety of transporting goods. Therefore, detecting the fault types of belt conveyors and performing corresponding repairs has become the research focus of belt conveyor management and maintenance.

[0003] Traditional fault detection methods for belt conveyors mainly rely on manual inspections. The operating status of the belt conveyor is monitored through the visual observation results and manual measurement results of the operators, and the fault types of the belt conveyor are judged manually. It is impossible to accurately determine the degree of damage of the belt conveyor, and the inspection results are affected by human factors. It is impossible to accurately judge the fault types of the belt conveyor, which easily leads to misjudgment or missed judgment of faults. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a fault detection method, which can solve the technical problems that the existing technology cannot determine the degree of damage of the belt conveyor and the accuracy of judging the fault types of the belt conveyor is relatively low.

[0005] In order to achieve the above purpose, the technical solutions adopted in the embodiments of the present invention are as follows:

[0006] In a first aspect, an embodiment of the present invention provides a fault detection method, including: acquiring a belt image of a conveying device, and identifying each damaged area in the belt image;

[0007] Segmenting and extracting each of the damaged areas from the belt image to obtain a plurality of segmented images;

[0008] Determining the degree of damage and fault type of the conveying device based on the damage characteristics of the segmented images; wherein the damage characteristics include the damaged area and / or the number of damages.

[0009] Further, an embodiment of the present invention provides a first possible implementation manner of the first aspect, wherein the damage characteristics include the damaged area and the number of damages, the degree of damage includes the overall degree of damage, and the step of determining the degree of damage and fault type of the conveying device based on the damage characteristics of the segmented images includes:

[0010] Determine the total damaged area and total damaged quantity of the belt based on the damaged area and damaged quantity in each of the segmented images;

[0011] Determine the overall damaged degree of the conveying device based on the total damaged area and the total damaged quantity;

[0012] If the overall damaged degree is greater than the first type of degree threshold, determine that the fault type of the conveying device is an overall physical damage type fault.

[0013] Furthermore, the embodiment of the present invention provides a second possible implementation manner of the first aspect, wherein the step of determining the overall damaged degree of the conveying device based on the total damaged area and the total damaged quantity includes:

[0014] Calculate the ratio of the total damaged quantity to the preset damaged quantity to obtain a quantity ratio;

[0015] Calculate the ratio of the total damaged area to the preset damaged area to obtain an area ratio;

[0016] Perform a summation calculation on the quantity ratio and the area ratio to obtain the overall damaged degree of the conveying device.

[0017] Furthermore, the embodiment of the present invention provides a third possible implementation manner of the first aspect, wherein the damaged feature includes the damaged quantity, the damage degree includes the segmented damage degree, and the step of determining the damage degree and fault type of the conveying device based on the damaged features of the segmented images includes:

[0018] Determine the belt width corresponding to each of the segmented images;

[0019] Determine the segmented damage degree of the conveying device based on the damaged quantity in each of the segmented images and the corresponding belt width;

[0020] If the segmented damage degree is greater than the second type of degree threshold, determine that the conveying device has a segmented physical damage type fault.

[0021] Furthermore, the embodiment of the present invention provides a fourth possible implementation manner of the first aspect, wherein the step of determining the segmented damage degree of the conveying device based on the damaged quantity in each of the segmented images and the corresponding belt width includes:

[0022] Calculate the ratio of the damaged quantity in each of the segmented images to the corresponding belt width to obtain the quantity-width ratio corresponding to each of the segmented images;

[0023] Perform a summation calculation on the quantity-width ratios corresponding to each of the segmented images to obtain the segmented damage degree of the conveying device.

[0024] Further, an embodiment of the present invention provides a fifth possible implementation manner of the first aspect, where the damage feature includes the damaged area, the damage degree includes the overall failure risk degree, and the step of determining the damage degree and failure type of the conveying device based on the damage features of the segmented images includes:

[0025] Determine the average damaged area of the segmented images based on the number of the segmented images and the damaged areas in each of the segmented images; determine the damage ratio corresponding to each of the segmented images based on the average damaged area and the damaged areas in each of the segmented images;

[0026] Determine the overall failure risk degree of the conveying device based on the damage ratios corresponding to each of the segmented images;

[0027] If the overall failure risk degree is greater than the third type of degree threshold, determine that the conveying device has an overall performance degradation type of failure.

[0028] Further, an embodiment of the present invention provides a sixth possible implementation manner of the first aspect, where the step of determining the average damaged area of the segmented images based on the number of the segmented images and the damaged areas in each of the segmented images; and determining the damage ratio corresponding to each of the segmented images based on the average damaged area and the damaged areas in each of the segmented images includes:

[0029] Perform a summation calculation on the damaged areas in each of the segmented images to obtain the total damaged area;

[0030] Calculate the ratio of the total damaged area to the number of the segmented images to obtain the average damaged area;

[0031] Calculate the ratio of the damaged area in each of the segmented images to the average damaged area to obtain the damage ratio corresponding to each of the segmented images.

[0032] Further, an embodiment of the present invention provides a seventh possible implementation manner of the first aspect, where the step of determining the overall failure risk degree of the conveying device based on the damage ratios corresponding to each of the segmented images includes:

[0033] Perform a summation calculation on the damage ratios corresponding to each of the segmented images to obtain the total damage ratio;

[0034] Calculate the ratio of the total damage ratio to the number of the segmented images to obtain the average damage ratio;

[0035] Perform a variance calculation on the damage ratios corresponding to each of the segmented images, the number of the segmented images, and the average damage ratio to obtain the damage ratio variance;

[0036] Sum the average breakage ratio and the breakage ratio variance to obtain the overall failure risk level of the conveying equipment.

[0037] Further, an eighth possible implementation manner of the first aspect is provided in an embodiment of the present invention, where the belt image includes a temperature image of the belt. The step of acquiring the belt image of the conveying equipment and identifying each breakage area in the belt image includes:

[0038] Acquire the temperature image of the belt in the conveying equipment;

[0039] Input the temperature image into the trained breakage area recognition model to identify abnormal temperature areas; determine an abnormal temperature list of the abnormal temperature areas based on the temperature values of the abnormal temperature areas; divide the abnormal temperature list into multiple segmented temperature lists based on a preset temperature segmentation interval; determine a breakage temperature list based on a preset temperature determination threshold, a preset temperature determination parameter, and each segmented temperature list; determine each breakage area in the temperature image based on the breakage temperatures in the breakage temperature list; where the breakage area recognition model is trained by a temperature image training data set marked with belt breakage areas.

[0040] Further, a ninth possible implementation manner of the first aspect is provided in an embodiment of the present invention, where the step of acquiring the belt image of the conveying equipment and identifying each breakage area in the belt image includes:

[0041] Acquire the belt image of the conveying equipment;

[0042] Input the belt image into the trained breakage area recognition model to identify each breakage area in the belt image; where the breakage area recognition model is trained by a belt image training data set marked with breakage areas.

[0043] An embodiment of the present invention provides a fault detection method, which includes: obtaining a belt image of a conveying device, and identifying each damaged area in the belt image; segmenting and extracting each of the damaged areas from the belt image to obtain a plurality of segmented images; determining the damage degree and fault type of the conveying device based on the damage features of the segmented images; wherein, the damage features include the damaged area and / or the damaged quantity. By identifying the damaged areas in the belt image, accurately positioning the damaged positions of the belt, extracting the segmented images, and determining the damage degree and fault type of the conveying device based on the damage features of the segmented images, the present invention realizes the automatic detection of the damage degree and fault type of the conveying device. Different from the prior art that relies on manual identification of the fault type of the conveying device, this method obtains the damage features of the belt through the belt image, and finally determines the damage degree and fault type of the conveying device, avoiding the influence of human factors, improving the accuracy of judging the fault type of the conveying device, and avoiding the occurrence of misjudgment or missed judgment.

[0044] Other features and advantages of the embodiments of the present invention will be described in the subsequent description, or some features and advantages can be inferred from the description or determined without doubt, or can be known by implementing the above technologies of the embodiments of the present invention.

[0045] To make the above objects, features and advantages of the present invention more obvious and understandable, the following specifically enumerates preferred embodiments and, in conjunction with the accompanying drawings, makes the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0047] Figure 1 The flowchart of a fault detection method provided by an embodiment of the present invention is shown;

[0048] Figure 2 The flowchart of a method for identifying damaged areas based on a temperature image provided by an embodiment of the present invention is shown;

[0049] Figure 3 The flowchart of a method for screening damaged areas in a temperature image by an abnormal analysis unit provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be described below with reference to the accompanying drawings. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention.

[0051] At present, belt conveyors are the main equipment for material transportation. With the long-term use of belt conveyors, damages occur to the belt conveyors, and it is inevitable that problems affecting the safety of the equipment and the continuity of the transported goods arise. Therefore, fault detection methods are often used to detect the fault types of belt conveyors for corresponding repairs. The fault detection methods in the prior art usually rely on manual inspections to determine the fault types of belt conveyors by humans. This easily leads to the neglect of potential problems of belt conveyors, and it is impossible to determine the degree of damage of belt conveyors. Moreover, the fault types of belt conveyors detected manually are easily affected by human factors. When manual inspection is used as a fault detection method, it is impossible to accurately judge the fault types of belt conveyors, and problems such as misjudgment or missed judgment are likely to occur. To improve the above problems, the embodiments of the present invention provide a fault detection method, which can be applied to accurately judge the degree of damage and fault types of belt conveyors, and avoid the occurrence of misjudgment or missed judgment. The following provides a detailed introduction to the embodiments of the present invention.

[0052] This embodiment provides a fault detection method, which can be applied to electronic devices such as computers. Refer to Figure 1 the flow schematic diagram of a fault detection method shown in the following figure. This method mainly includes the following steps:

[0053] Step S101, obtain the belt image of the conveying equipment, and identify each damaged area in the belt image;

[0054] The above-mentioned conveying equipment includes a belt conveyor. The belt image of the belt conveyor is obtained through an image acquisition system. The image acquisition system includes a camera and a light source. Usually, the camera is set above the frame of the belt conveyor to take pictures of the belt area of the belt conveyor to obtain the belt image of the belt conveyor. And when taking pictures through the camera, a light source is usually also set around the camera to make the belt image of the belt conveyor obtained by the camera clearer;

[0055] Obtain the damage forms and degrees of damaged areas that are common and within an acceptable range during the normal operation of the belt conveyor as training samples, and perform statistical analysis on a large number of training sample data based on a damage parameter threshold acquisition model to obtain a set of benchmark damage parameter threshold sets. There are various benchmark damage parameter thresholds in the benchmark damage parameter threshold set. Specifically, it includes: the minimum area threshold of the damaged area the minimum length threshold the minimum width threshold the damage shape feature threshold Preset damage quantity corresponding to the belt Preset damage area Preset area quantity ratio Preset segmentation damage quantity corresponding to the segmented image Preset segmentation damage area Preset temperature threshold P corresponding to the belt x10 ;

[0056] Among them, various benchmark damage parameter thresholds in the benchmark damage parameter threshold set are as follows:

[0057]

[0058] Among them, is the a-th benchmark damage parameter threshold, a = 1, 2,..., 10, the total number of training samples is T, p ai is the a-th benchmark damage parameter threshold in the i-th training sample, ω ai is the weight corresponding to the a-th benchmark damage parameter threshold in the i-th training sample;

[0059] Based on the minimum area threshold Minimum length threshold Minimum width threshold Damage shape feature threshold of the damaged area in the benchmark damage parameter threshold set, the identified damaged area can be further screened to accurately define the damaged area, prevent misjudgment, use the minimum area threshold, minimum length threshold and minimum width threshold to filter out tiny interference areas, and the damage shape feature threshold helps to distinguish the pseudo-damaged areas in the damaged area, improving the accuracy of the identified damaged area and the precision in subsequent fault type detection;

[0060] The benchmark damage parameter threshold set reflects the characteristics of the defect areas of the training samples, enabling better distinction of the damaged areas when identifying the damaged areas in the belt image. When the benchmark damage parameter threshold set is applied to any segmented image and used as a reference benchmark, it can be used to judge the possibility of problems in each segmented image. For example, if the damage features in a certain segmented image are significantly higher than the corresponding benchmark damage parameter thresholds in the benchmark damage parameter threshold set, then this segmented image is more likely to have a damage problem.

[0061] Step S103, segment and extract each damaged area from the belt image to obtain multiple segmented images;

[0062] In the above steps, when segmenting and extracting the belt image, the damaged area is obtained relying on the pixel gray value, color, texture features, edge contour and temperature in the belt image. For example, when the pixel gray value, color and temperature of a certain area in the belt image are significantly different from those of the surrounding area, it is determined as the damaged area; or when the texture features of the belt image show irregular and broken conditions, the area is determined as the damaged area; or when the edge contour of the belt image is discontinuous or has sharp turns, etc., the area is determined as the damaged area; the determined damaged areas in the belt image are segmented and extracted from the belt image to obtain multiple segmented images, laying a foundation for accurately obtaining the damage degree and judging the fault type later.

[0063] Step S105, determining the damage degree and fault type of the conveying equipment based on the damage features of the segmented images; wherein, the damage features include the damaged area and / or the number of damages.

[0064] Establish a two-dimensional coordinate axis based on the belt image, identify the number of damaged areas in each segmented image, the central coordinates of each damaged area in the two-dimensional coordinate axis and the damage type of each damaged area based on the damage feature recognition model, and establish a damage identification set:

[0065]

[0066] wherein, P k is the damage identification set of the k-th segmented image, n k is the number of damages of all damaged areas in the k-th segmented image, c i is the central coordinate of the i-th damaged area in the two-dimensional coordinate axis of the k-th segmented image, and t i is the damage type of the i-th damaged area in the k-th segmented image.

[0067] The damage types of the above damaged areas include: (1) Surface scratch: Usually the area where a sharp object contacts the belt to produce linear marks; (2) Wear: Usually the area of material loss on the belt surface formed by the long-term friction between the belt and the idler or other components; (3) Tear: Usually the area with a large crack formed when the belt is subjected to a large tensile force or is involved in a rotating component; (4) Perforation: Usually the area where the belt has circular or irregular holes caused by being penetrated by foreign objects; (5) Spalling: Usually the area where the belt surface material falls off.

[0068] The above damage identification set contains the identification information of all damaged areas in each segmented image, determines the positions of each damaged area in each segmented image based on the central coordinates, and clarifies the damage types of each damaged area, which helps to obtain the damage features of each segmented image, that is, the damaged area of each damaged area and the number of damaged areas in each segmented image.

[0069] The damaged area and the number of damages of the segmented image can be used to determine the overall damage degree of the conveying equipment, and can be used to judge whether the conveying equipment has an overall physical damage type fault; the number of damages of the segmented image can be used to determine the segmented damage degree of the conveying equipment, and can be used to judge whether the conveying equipment has a segmented physical damage type fault; the damaged area of the segmented image can be used to determine the overall fault risk degree of the conveying equipment, and can be used to judge whether the conveying equipment has an overall performance degradation type fault.

[0070] The above method provided by the embodiments of the present invention realizes the automatic detection of the damage degree and fault type of the conveying equipment by obtaining the belt image of the conveying equipment (including the belt conveyor), identifying the damaged area in the belt image, accurately positioning the damaged position of the belt, extracting the segmented image, and determining the damage degree and fault type of the conveying equipment based on the damage characteristics of the segmented image. Different from the prior art that relies on manual identification of the fault type of the conveying equipment, this method obtains the damage characteristics of the belt through the belt image, and finally determines the damage degree and fault type of the conveying equipment, avoiding the subjective influence of human factors, improving the accuracy of judging the fault type of the conveying equipment, and avoiding the occurrence of misjudgment or missed judgment.

[0071] In one embodiment, the damage characteristics provided in this embodiment include the damaged area and the number of damages, and the damage degree includes the overall damage degree. The specific implementation manner of determining the damage degree and fault type of the conveying equipment based on the damage characteristics of the segmented image includes:

[0072] Determine the total damaged area and the total number of damages of the belt according to the damaged area and the number of damages in each segmented image;

[0073] When judging the overall damage degree of the conveying equipment, the damaged area and the number of damages of each segmented image are required. The damaged area and the number of damages in each segmented image are respectively summed to obtain the total damaged area S and the total number of damages N of the belt:

[0074]

[0075] Among them, S is the total damaged area of the belt, N is the total number of damages of the belt, M is the total number of segmented images, A k is the total damaged area of all damaged areas in the k-th segmented image, n k is the total number of damages of all damaged areas in the k-th segmented image.

[0076] Determine the overall damage degree of the conveying equipment based on the total damaged area and the total number of damages;

[0077] Obtain the area-number ratio based on the total damaged area and the total number of damages:

[0078]

[0079] Among them, P is the average damage severity represented by the unit damaged area, that is, the area quantity ratio, S is the total damaged area, and N is the total number of damages. Obtaining the area quantity ratio based on the total damaged area and the total number of damages can quantify the severity of each damaged area, facilitating the subsequent correction of the number of damages;

[0080] Compare the area quantity ratio P with the preset area quantity ratio to achieve the correction of the total number of damages N:

[0081]

[0082] Among them, N new is the corrected total number of damages, is the total error correction coefficient, which represents the relationship between the area quantity ratio P and the preset area quantity ratio . Appropriately increase or decrease the total number of damages according to the total error correction coefficient, and appropriately adjust the correction degree of the total number of damages (i.e., the increased or decreased quantity range) according to the deviation degree between the area quantity ratio P and the preset area quantity ratio . By obtaining the corrected total number of damages, optimize the total number of damages, avoid missing smaller damaged areas, so as to ensure that the corrected total number of damages can improve the accuracy of the subsequent obtained damage degree and fault type;

[0083] Determine the overall damage degree of the conveying equipment based on the total damaged area and the corrected total number of damages.

[0084] If the overall damage degree is greater than the first type of degree threshold, determine that the fault type of the conveying equipment is an overall physical damage type fault;

[0085] The above-mentioned first type of degree threshold is a preset overall damage degree threshold. When the overall damage degree is greater than the first type of degree threshold, it proves that the conveying equipment is an overall physical damage type fault.

[0086] In one embodiment, the specific implementation method provided in this embodiment for determining the overall damage degree of the conveying equipment based on the total damaged area and the total number of damages includes:

[0087] Calculate the ratio of the total number of damages to the preset number of damages to obtain the quantity ratio;

[0088] Obtain the corrected total number of damages, calculate the ratio of the corrected total number of damages to the preset number of damages, and get:

[0089]

[0090] Among them, N ratio is the quantity ratio, Nnew is the total corrected damaged quantity, is the preset damaged quantity.

[0091] Calculate the ratio of the total damaged area to the preset damaged area to obtain the area ratio;

[0092] Obtain the total corrected damaged area, calculate the ratio of the total corrected damaged area to the preset damaged area, and obtain:

[0093]

[0094] where S ratio is the area ratio, S is the total corrected damaged area, is the preset damaged area.

[0095] Sum the quantity ratio and the area ratio to obtain the overall damaged degree of the conveying equipment;

[0096] The overall damaged degree of the conveying equipment obtained is:

[0097]

[0098] where F belt1 is the overall damaged degree of the conveying equipment, ω N is the preset weight of the quantity ratio, ω S is the preset weight of the area ratio;

[0099] The above calculation of the overall damaged degree of the conveying equipment is carried out by comparing the corrected damaged quantity and the total damaged area with the corresponding preset damaged quantity and preset damaged area respectively, and performing weighted calculation according to the results of the quantity ratio and the area ratio (or calculating according to specific calculation rules) to finally obtain the overall damaged degree of the conveying equipment. For example, when the corrected damaged quantity is large and the total damaged area is large, the obtained quantity ratio and area ratio are large, and when the obtained overall damaged degree of the conveying equipment is large and greater than the first type of degree threshold, it is determined that the conveying equipment has an overall physical damage type fault and a serious fault, and the conveying equipment is repaired in time.

[0100] In one embodiment, the damaged features provided in this embodiment include the damaged quantity, and the damage degree includes the segmented damage degree. The specific implementation manner of determining the damage degree and fault type of the conveying equipment based on the damaged features of the segmented images includes:

[0101] Determine the belt width corresponding to each segmented image;

[0102] When judging the segmented damage degree of the conveying equipment, the damaged quantity of each segmented image needs to be used, and the belt width corresponding to each segmented image is determined. Among them, the belt width corresponding to the k-th segmented image is Bk 。

[0103] Determine the degree of segmentation damage of the conveying equipment based on the number of damages in each segmented image and the corresponding belt width;

[0104] Obtain the number of damages in each segmented image, where the number of damages corresponding to the k-th segmented image is n k , and compare the number of damages in each segmented image with the preset number of segmented damages corresponding to the segmented image to correct the number of damages in each segmented image:

[0105]

[0106] where N′ k is the corrected number of damages in the k-th segmented image, is the segmentation error correction coefficient, indicating the relationship between the number of damages in the k-th segmented image and the preset number of segmented damages corresponding to the segmented image ;

[0107] Appropriately increase or decrease the number of damages in each segmented image according to the segmentation error correction coefficient. For example, when the number of damages in the k-th segmented image is significantly lower than the number of segmented damages , the number of damages in the segmented image may be reduced. Even because the number of damages in this segmented image is small, the degree of damage of this segmented image may be regarded as within the normal fluctuation range, or this segmented image may be regarded as an unimportant segmented image caused by detection errors, so this segmented image is excluded; and when the number of damages in the k-th segmented image exceeds the number of segmented damages , the number of damages corresponding to this segmented image will be adjusted to a more reasonable range; adjust the number of damages in each segmented image through the segmentation error correction coefficient, so that the number of damages in each segmented image more conforms to the actual fault situation, thereby reducing excessive maintenance caused by misjudgment and ensuring the accuracy and reliability when predicting the fault type of the conveying equipment later;

[0108] Determine the degree of segmentation damage of the conveying equipment based on the corrected number of damages in each segmented image and the corresponding belt width.

[0109] If the degree of segmentation damage is greater than the second type of degree threshold, determine that the conveying equipment has a segmentation physical damage type fault;

[0110] The above-mentioned second type of degree threshold is a preset segmentation damage degree threshold. When the degree of segmentation damage is greater than the first type of degree threshold, it is proved that the conveying equipment has a segmentation physical damage type fault.

[0111] In one embodiment, the specific implementation of determining the segmentation damage degree of the conveying device based on the number of damages in each segmented image and the corresponding belt width includes:

[0112] Calculate the ratio of the number of damages in each segmented image to the corresponding belt width to obtain the number-width ratio corresponding to each segmented image;

[0113] Obtain the corrected number of damages in each segmented image, calculate the ratio of the corrected number of damages in each segmented image to the corresponding belt width, and obtain the number-width ratio corresponding to the segmented image:

[0114]

[0115] where D k is the number-width ratio corresponding to the k-th segmented image, N′ k is the corrected number of damages in the k-th segmented image, and B k is the belt width corresponding to the k-th segmented image.

[0116] Perform a summation calculation on the number-width ratios corresponding to each segmented image to obtain the segmentation damage degree of the conveying device;

[0117] Perform a summation calculation on the number-width ratios corresponding to each segmented image to obtain the segmentation damage degree of the conveying device:

[0118]

[0119] where the belt of the conveying device corresponds to M segmented images, and F belt2 is the segmentation damage degree of the conveying device, and w k is the weight corresponding to the k-th segmented image;

[0120] By performing a weighted summation on the number-width ratios (i.e., corrected damage ratios) of all segmented images, the obtained F belt2 can more comprehensively reflect the segmentation damage degree of the conveying device. This segmentation damage degree takes into account the number of damages in each segmented image and the corresponding belt width, and measures the segmentation damage degree corresponding to the overall belt.

[0121] In one embodiment, assume that the belt of a certain conveying device includes three segmented images, and the corrected number of damages corresponding to the three segmented images are N1′ = 5, N2′ = 3, N3′ = 4 respectively. The bandwidths corresponding to the three segmented images are B1 = 10, B2 = 8, B3 = 12 respectively, and the weights of the bandwidths corresponding to the three segmented images are ω1 = 0.4, ω2 = 0.3, ω3 = 0.3 respectively. The segmentation damage degree corresponding to this conveying device is:

[0122]

[0123] Set the second - type degree threshold to 0.4. Since the corresponding segmentation damage degree of the conveying device is greater than the second - type degree threshold, it is determined that the conveying device has a physical damage failure due to segmentation.

[0124] In one embodiment, the damage features provided in this embodiment include the damaged area, and the damage degree includes the overall failure risk degree. The specific implementation manner for determining the damage degree and failure type of the conveying device based on the damage features of the segmented images is as follows:

[0125] When judging the overall failure risk degree of the conveying device, the damaged areas of the segmented images are required. Determine the average damaged area of the segmented images based on the number of segmented images and the damaged areas in each segmented image; determine the damage ratio corresponding to each segmented image based on the average damaged area and the damaged areas in each segmented image;

[0126] Calculating the damage ratio corresponding to each segmented image can effectively measure the corresponding damage condition (i.e., the damage degree) of each segmented image in the belt. For example, when the damage ratio is close to 1, it indicates that the damage condition of this segmented image is close to the average level and is at a relatively normal damage degree; when the damage ratio is greater than 1, it indicates that the damage condition of this segmented image is higher than the average level and is at a higher damage degree, and this segmented image needs to be focused on.

[0127] Determine the overall failure risk degree of the conveying device based on the damage ratios corresponding to each segmented image;

[0128] If the overall failure risk degree is greater than the third - type degree threshold, it is determined that the conveying device has an overall performance degradation failure;

[0129] The above - mentioned third - type degree threshold is a pre - set overall failure risk degree threshold. When the overall failure risk degree is greater than the third - type degree threshold, it proves that the conveying device has an overall performance degradation failure.

[0130] In one embodiment, the specific implementation manner for determining the average damaged area of the segmented images based on the number of segmented images and the damaged areas in each segmented image; and determining the damage ratio corresponding to each segmented image based on the average damaged area and the damaged areas in each segmented image is as follows:

[0131] Sum up the damaged areas in each segmented image to obtain the total damaged area;

[0132] The above - mentioned summing up the damaged areas in each segmented image gives the total damaged area S of the belt;

[0133] Calculate the ratio of the total damaged area to the number of segmented images to obtain the average damaged area;

[0134] Set the number of segmented images as M, calculate the ratio of the total damaged area to the number of segmented images to obtain the average damaged area:

[0135]

[0136] Among them, is the average damaged area of the segmented images.

[0137] Calculate the ratio of the damaged area in each segmented image to the average damaged area to obtain the damage ratio corresponding to each segmented image;

[0138] Calculate the damaged area in each segmented image:

[0139]

[0140] Among them, A k is the total damaged area in the k-th segmented image, n k is the number of damages in the k-th segmented image, a i is the damaged area of the i-th damaged region in the k-th segmented image;

[0141] The total damaged area of each segmented image obtained above can be used to measure the overall defect degree of each segmented image, further quantify the damaged situation of each segmented region. When the total damaged area of the segmented image exceeds the preset segmented damaged area At this time, the area corresponding to the segmented image in the belt is identified as a potential fault area (i.e., an area with a high fault risk). For example, if a segmented image contains 10 damaged regions and the total damaged area in this segmented image reaches 100 square millimeters, while the preset segmented damaged area is 80 square millimeters, then this segmented image is identified as an area with a high fault risk and a warning signal is output to prompt the personnel to pay attention to this segmented image;

[0142] Calculate the ratio of the damaged area in each segmented image to the average damaged area to obtain the damage ratio corresponding to each segmented image:

[0143]

[0144] Among them, R k is the damage ratio corresponding to the k-th segmented image, which is used to represent the damaged situation of the k-th segmented image in the entire belt.

[0145] In one embodiment, the specific implementation manner provided in this embodiment for determining the overall fault risk degree of the conveying equipment based on the damage ratio corresponding to each segmented image includes:

[0146] Perform a summation calculation on the damage ratios corresponding to each segmented image to obtain the total damage ratio;

[0147] Sum up the damage ratios corresponding to each segmented image to obtain the total damage ratio:

[0148]

[0149] Among them, R all is the total damage ratio, and M is the number of segmented images.

[0150] Calculate the ratio of the total damage ratio to the number of segmented images to obtain the average damage ratio;

[0151] Calculate the ratio of the total damage ratio to the number of segmented images to obtain the average damage ratio:

[0152]

[0153] Among them, μ R is the average damage ratio, indicating the average damage degree of all segmented images.

[0154] Calculate the variance of the damage ratios, the number of segmented images, and the average damage ratio corresponding to each segmented image to obtain the damage ratio variance;

[0155] Calculate the variance of the damage ratios, the number of segmented images, and the average damage ratio corresponding to each segmented image to obtain the damage ratio variance:

[0156]

[0157] Among them, is the damage ratio variance, indicating the fluctuation of the damage degrees of all segmented images.

[0158] Sum up the average damage ratio and the damage ratio variance to obtain the overall failure risk degree of the conveying equipment;

[0159] Set the damage value proportion coefficient of the belt, multiply the sum of the average damage ratio and the damage ratio variance by the damage value proportion coefficient to obtain the overall failure risk degree of the conveying equipment:

[0160]

[0161] Among them, F belt3 is the overall failure risk degree of the conveying equipment, C is the damage value proportion coefficient, and different values can be set according to different usage scenarios of the conveying equipment;

[0162] By introducing the average damage ratio and the damage ratio variance, F belt3 can not only reflect the overall failure risk degree of the conveying equipment, but also reflect the fluctuation of the damage degrees of each segmented image. When the damage ratio variance is large, that is, the damage degrees among the segmented images are quite different, F belt3will be at a relatively high value, indicating that the damage degree of the belt of the conveying equipment is unstable and more intensive maintenance may be required.

[0163] In one embodiment, assume that the belt of a certain conveying equipment includes five segmented images, and the corresponding breakage ratios of the five segmented images are 0.8, 1.2, 1.0, 1.1, and 0.9 respectively. Then the average breakage ratio μ R ′≈1.0, and the breakage ratio variance Set the damage value proportionality coefficient to 1.5. Then the corresponding segmented damage degree of the conveying equipment is:

[0164] F belt3 = 1.5·(1 + 0.02) = 1.53;

[0165] Set the third type of degree threshold to 1.2. The corresponding overall failure risk degree of the conveying equipment is greater than the third type of degree threshold. It is determined that the conveying equipment is a failure of overall performance degradation, indicating that the overall failure risk degree of the belt of the conveying equipment is relatively high and the damage degrees of its segmented images vary greatly. The conveying equipment is a failure of overall performance degradation, and the maintenance frequency of the conveying equipment can be increased.

[0166] In one embodiment, the belt image provided in this embodiment includes the temperature image of the belt. The specific implementation manner of obtaining the belt image of the conveying equipment and identifying each damaged area in the belt image includes:

[0167] Obtain the temperature image of the belt in the conveying equipment;

[0168] Refer to Figure 2 As shown in the schematic flow diagram of identifying damaged areas based on temperature images, the image acquisition system further includes an infrared thermal imager, which is arranged above the frame of the conveying equipment (i.e., belt conveyor) and is used to monitor the temperature of the belt of the conveying equipment in real time and obtain the temperature image of the belt;

[0169] Input the temperature image into the trained damaged area recognition model to identify the abnormal temperature area;

[0170] The training steps of the above damaged area recognition model include: Mark the damaged areas of the conveying equipment corresponding to the temperature images of the belts of each conveying equipment to obtain a temperature image training data set; Input the temperature image training data set into the damaged area recognition model respectively to perform model training on the damaged area recognition model to obtain the trained damaged area recognition model;

[0171] Obtain the areas in the temperature image that have a large difference from the preset temperature threshold P xl0 The areas with uneven temperature distribution and the boundary areas where the temperature changes suddenly to obtain the abnormal temperature areas in the temperature image.

[0172] Determine the abnormal temperature list of the abnormal temperature area based on the temperature values of each abnormal temperature area;

[0173] Construct the abnormal temperature list of the abnormal temperature area based on the temperature values of each abnormal temperature area (i.e., suspected temperature area):

[0174] T ist ={T i |T i =T suspect,i ,T i+1 =T suspect,i+1}};

[0175] Among them, T ist is the abnormal temperature list of the abnormal temperature area, which is convenient for observing the abnormal temperatures in each abnormal temperature area and facilitating the subsequent screening of the abnormal temperature area. T suspect,i is the temperature value of the i-th abnormal temperature area, and T suspect,i+1 is the temperature value of the (i + 1)-th abnormal temperature area. Based on the temperature fluctuation range during normal operation of the device and combined with a certain temperature safety margin, set the preset temperature screening threshold, and screen the abnormal temperature areas corresponding to each abnormal temperature in the abnormal temperature list based on the preset temperature screening threshold and temperature change.

[0176] Divide the abnormal temperature list into multiple segmented temperature lists based on the preset temperature segmentation interval;

[0177] Determine the preset temperature segmentation interval according to the size of the abnormal temperature area, the gradient of temperature change, the structural characteristics of the conveying device, etc.;

[0178] Divide the abnormal temperature list into multiple segmented temperature lists based on the preset temperature segmentation interval:

[0179] T split,v ={T ist |T ist,v =segmentation interval};

[0180] Among them, T split,v is the segmented temperature list, and T ist,v is the abnormal temperature list divided by the segmentation interval; Dividing the temperature list through the preset temperature segmentation interval can divide the temperature data of the temperature list into different sub-areas to obtain multiple segmented temperature lists, so as to conduct temperature difference analysis in a more detailed temperature area.

[0181] Determine the damaged temperature list based on the preset temperature determination threshold, preset temperature determination parameter, and each segmented temperature list;

[0182] Set the preset temperature determination threshold and preset temperature determination parameter through statistical analysis of a large amount of temperature data when the conveying device is in normal operation;

[0183] Determine the list of damaged temperatures based on the preset temperature determination threshold, preset temperature determination parameter, and each segmented temperature list:

[0184] T defect ={T split,w |P judge,w ≥θ};

[0185] Among them, T defect is the list of damaged temperatures, T split,w is the w-th segmented temperature list, P judge,w is the preset temperature determination parameter of the w-th segmented temperature list, and θ is the preset temperature determination threshold;

[0186] By obtaining the list of damaged temperatures and based on the damaged temperatures included in the list of damaged temperatures, the position and range of the damaged area can be more accurately reflected.

[0187] Determine each damaged area in the temperature image based on the damaged temperatures in the list of damaged temperatures; among them, the damaged area recognition model is trained by a temperature image training data set marked with the belt damaged area;

[0188] Obtain the damaged area determination parameter based on the damaged temperatures in the list of damaged temperatures:

[0189]

[0190] Among them, P defect is the damaged area determination parameter, n is the number of damaged temperatures included in the list of damaged temperatures, T defect,v is the v-th temperature value in the list of damaged temperatures, and T defect,v+1 is the (v + 1)-th temperature value in the list of damaged temperatures;

[0191] Set the preset temperature difference threshold for the damaged area, and determine the area corresponding to the damaged area determination parameter that is not within the preset temperature difference threshold as the final damaged area, and calibrate the exact position of the damaged area in the temperature image and the temperature difference of the damaged area. For example, in a certain list of damaged temperatures, the v-th temperature value is 70 degrees, the (v + 1)-th temperature value is 80 degrees, then the temperature difference of the damaged area is 10 degrees, and the preset temperature difference threshold is 8 degrees. Since this temperature difference value exceeds the preset temperature difference threshold, determine that this area is the damaged area.

[0192] In the embodiment of the present invention, referring to the flow schematic diagram of screening the damaged area in the temperature image by an anomaly analysis unit as shown in Figure 3 , after determining the damaged area through the temperature image, the anomaly analysis unit can further analyze and screen the damaged area by obtaining the temperature rise parameter of the damaged area;

[0193] Obtain temperature rise area, temperature rise duration, temperature rise cycle and other temperature rise parameters for each damaged area, and obtain the average temperature rise area, average temperature rise duration and average temperature rise cycle corresponding to all damaged areas. These two sets of data will be used to calculate the difference degree between the temperature rise parameters of each damaged area and the average temperature rise parameters;

[0194] Calculate the first difference degree as:

[0195]

[0196] where D1 is the first difference degree, representing the relative difference between the temperature rise duration of the damaged area and the average temperature rise duration, reflecting the abnormal performance of the damaged area in terms of time, t defect is the temperature rise duration of each damaged area, and t avg is the average temperature rise duration corresponding to the damaged area;

[0197] Calculate the second difference degree as:

[0198]

[0199] where D2 is the second difference degree, representing the difference degree between the temperature rise area of each damaged area and the average temperature rise area, reflecting whether the temperature rise area of the damaged area is significantly higher than the normal level, A defect is the temperature rise area of each damaged area, and A avg is the average temperature rise area corresponding to the damaged area;

[0200] By calculating the product of the first difference degree and the second difference degree, obtain the first product to comprehensively evaluate the abnormality degree of the damaged area:

[0201] Y1 = D1 × D2;

[0202] where Y1 is the first product, reflecting the comprehensive abnormality degree of the damaged area when in the temperature rise state. The larger its value, the higher the probability of damage to the damaged area;

[0203] Set the temperature rise coefficient Z to finally obtain the temperature rise parameters of the damaged area:

[0204] Y rise = Y1 × Z;

[0205] where Y rise is the temperature rise parameter of damage. By calculating the product of the first product and the temperature rise coefficient, the quantitative analysis of the temperature rise behavior of the damaged area is realized.

[0206] Set the threshold value Y of the temperature rise parameter of damage threshold, compare the calculated damaged temperature rise parameter with the damaged temperature rise parameter threshold. When the damaged temperature rise parameter is greater than the damaged temperature rise parameter threshold, mark the damaged area as having a high degree of damage for subsequent maintenance and processing of the damaged area. For example, assume that the temperature rise area of a damaged area is 150 square centimeters, the corresponding average temperature rise area is 100 square centimeters, the temperature rise duration is 30 seconds, the corresponding average duration is 20 seconds, the temperature rise coefficient is 1.5. According to the above formula, the first difference degree is calculated to be 0.5, the second difference degree is 0.5, the first product is 0.25, the damaged temperature rise parameter is 0.375, and the set damaged temperature rise parameter threshold is 0.5. Then, this damaged area will not be marked as an area with a high degree of damage, which can effectively identify the damage degree of each damaged area, thereby improving the accuracy of maintenance.

[0207] In one embodiment, this embodiment provides a specific implementation method for obtaining the belt image of the conveying device and identifying each damaged area in the belt image, including:

[0208] Obtain the belt image of the conveying device;

[0209] Obtain the belt image of the conveying device through the camera in the image acquisition system.

[0210] Input the belt image into the trained damaged area recognition model to identify each damaged area in the belt image; wherein, the damaged area recognition model is trained by a belt image training data set marked with damaged areas.

[0211] The training steps of the above-mentioned damaged area recognition model include: marking the damaged areas of the conveying devices corresponding to the belt images of each conveying device to obtain a belt image training data set; respectively inputting the belt image training data set into the damaged area recognition model to perform model training on the damaged area recognition model to obtain the trained damaged area recognition model.

[0212] In the embodiment of the present invention, after detecting the damage degree and fault type of the conveying device through the fault detection method provided by the present invention, an alarm signal is sent in a timely manner to prompt personnel to repair the conveying device, realizing real-time maintenance of the conveying device.

[0213] On the basis of the foregoing embodiment, this embodiment provides an example of using the foregoing fault detection method to detect the fault type of a belt conveyor, which can be specifically executed with reference to the following steps:

[0214] Step S201: Obtain the belt image and the temperature image of the conveyor equipment through an image acquisition system, and identify the damaged areas in the belt image and the temperature image through a trained damaged area recognition model. Among them, the damaged area recognition model is constructed based on a deep learning algorithm and trained with a large number of belt image training datasets and temperature image training datasets.

[0215] Step S203: Segment and extract the damaged areas from the belt image and the temperature image to obtain multiple segmented images, and obtain the damaged features corresponding to each segmented image.

[0216] Step S205: Train a large number of normal damage situations of the conveyor equipment through an autoencoder to obtain a set of benchmark damage parameter threshold sets for subsequent use.

[0217] Step S207: Determine the overall damage degree of the conveyor equipment based on the benchmark damage parameter threshold, the total damaged area, and the total number of damages of the belt. If the overall damage degree is greater than the first type of degree threshold, determine that the fault type of the conveyor equipment is an overall physical damage type fault, send an alarm signal, locate the position of the damaged area, and prompt personnel to maintain the conveyor equipment.

[0218] Step S209: Determine the segmented damage degree of the conveyor equipment based on the benchmark damage parameter threshold, the number of damages in each segmented image, and the corresponding belt width. If the segmented damage degree is greater than the second type of degree threshold, determine that the conveyor equipment is a segmented physical damage type fault, send an alarm signal, locate the position of the damaged area, and prompt personnel to maintain the conveyor equipment.

[0219] Step S211: Determine the overall fault risk degree of the conveyor equipment based on the damage ratio corresponding to each segmented image. If the overall fault risk degree is greater than the third type of degree threshold, determine that the conveyor equipment is an overall performance degradation type fault.

[0220] Step S213: If the conveyor equipment is any one or more of the above fault types, send an alarm signal, locate the position of the damaged area, and prompt personnel to maintain the conveyor equipment.

[0221] The above method provided by the embodiments of the present invention integrates image processing and infrared thermal imaging technologies through the high-resolution camera and infrared thermal imager set in the image acquisition system, can capture the belt image of the conveying equipment and the temperature image of the belt in real time, combines the deep learning algorithm, establishes a damage area recognition model, and can accurately identify the damage areas in the belt image and the temperature image of the belt through the damage area recognition model. It can also accurately identify the tiny damage areas and overheated damage areas, improving the accuracy of identifying the damage areas, ensuring that the faults of the conveying equipment can be detected in the initial stage, improving the fault detection accuracy and response speed, greatly reducing the production downtime caused by the faults of the conveying equipment, and improving the production efficiency;

[0222] During the detection process of the damage degree and fault type of the conveying equipment, through the powerful data processing ability of the deep learning algorithm for in-depth analysis, when it is detected that the damage degree of the conveying equipment exceeds the threshold and belongs to one or more fault types, the damage area of the conveying equipment can be located, a visual report can be generated, a fault warning can be given, prompting the personnel to maintain the conveying equipment, and a clear damage area position is provided as the basis for fault diagnosis, enabling the personnel to formulate a more effective maintenance plan, helping the enterprise to optimize resource allocation and reduce maintenance costs;

[0223] Through this fault detection method, the automatic detection of the damage degree and fault type of the conveying equipment can be realized, avoiding the subjective influence of human factors, improving the accuracy of judging the fault type of the conveying equipment, avoiding the occurrence of misjudgment or missed judgment, avoiding the dependence on manual inspection, thus reducing the labor cost. Through the real-time monitoring and real-time warning of the conveying equipment, it can effectively prevent the safety accidents caused by the faults of the conveying equipment, ensure the safety of the operators, maintain the stability of the conveying process, ensure the safety and efficiency of the conveying environment, avoid the subjective influence of human factors, and improve the accuracy of judging the fault type of the conveying equipment, avoiding the occurrence of misjudgment or missed judgment.

[0224] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus cannot be understood as a limitation of the present invention. In addition, the terms "first", "second", "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0225] Finally, it should be noted that the above embodiments are only specific embodiments of the present invention, used to illustrate the technical solutions of the present invention, rather than limiting it. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that any person skilled in the technical field of the present invention can still modify the technical solutions described in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A fault detection method, characterized in that, Including: Obtain the belt image of the conveying device and identify each damaged area in the belt image; Segment and extract each of the damaged areas from the belt image to obtain a plurality of segmented images; Determine the damage degree and fault type of the conveying device based on the damage characteristics of the segmented images; wherein, the damage characteristics include the damaged area and / or the number of damages.

2. The fault detection method according to claim 1, characterized in that The damage characteristics include the damaged area and the number of damages, the damage degree includes the overall damage degree, and the step of determining the damage degree and fault type of the conveying device based on the damage characteristics of the segmented images includes: Determine the total damaged area and the total number of damages of the belt according to the damaged area and the number of damages in each of the segmented images; Determine the overall damage degree of the conveying device based on the total damaged area and the total number of damages; If the overall damage degree is greater than the first type of degree threshold, determine that the fault type of the conveying device is an overall physical damage type fault.

3. The fault detection method according to claim 2, wherein The step of determining the overall damage degree of the conveying device based on the total damaged area and the total number of damages includes: Calculate the ratio of the total number of damages to the preset number of damages to obtain a quantity ratio; Calculate the ratio of the total damaged area to the preset damaged area to obtain an area ratio; Perform a summation calculation on the quantity ratio and the area ratio to obtain the overall damage degree of the conveying device.

4. The fault detection method according to claim 1, wherein The damage characteristics include the number of damages, the damage degree includes the segmented damage degree, and the step of determining the damage degree and fault type of the conveying device based on the damage characteristics of the segmented images includes: Determine the belt width corresponding to each of the segmented images; Determine the segmented damage degree of the conveying device based on the number of damages in each of the segmented images and the corresponding belt width; If the segmented damage degree is greater than the second type of degree threshold, determine that the conveying device has a segmented physical damage type fault.

5. The fault detection method according to claim 4, wherein The step of determining the segmented damage degree of the conveying device based on the number of damages in each of the segmented images and the corresponding belt width includes: Calculate the ratio of the number of damages in each of the segmented images to the corresponding belt width to obtain the quantity-width ratio corresponding to each of the segmented images; Perform a summation calculation on the quantity-width ratios corresponding to each of the segmented images to obtain the segmented damage degree of the conveying device.

6. The fault detection method according to claim 1, wherein The damage characteristics include the damaged area, the damage degree includes the overall fault risk degree, and the step of determining the damage degree and fault type of the conveying device based on the damage characteristics of the segmented images includes: Determine the average damaged area of the segmented images based on the number of the segmented images and the damaged area in each of the segmented images; determine the damage ratio corresponding to each of the segmented images based on the average damaged area and the damaged area in each of the segmented images; Determine the overall fault risk degree of the conveying device based on the damage ratios corresponding to each of the segmented images; If the overall fault risk degree is greater than the third type of degree threshold, determine that the conveying device has an overall performance degradation type fault.

7. The fault detection method according to claim 6, wherein The step of determining the average damage area of the segmented images based on the number of the segmented images and the damage areas in each of the segmented images, and determining the damage ratio corresponding to each of the segmented images based on the average damage area and the damage areas in each of the segmented images includes: Performing a summation calculation on the damage areas in each of the segmented images to obtain the total damage area; Calculating the ratio of the total damage area to the number of the segmented images to obtain the average damage area; Calculating the ratio of the damage area in each of the segmented images to the average damage area to obtain the damage ratio corresponding to each of the segmented images.

8. The fault detection method according to claim 6, wherein The step of determining the overall failure risk level of the conveying device based on the damage ratios corresponding to each of the segmented images includes: Performing a summation calculation on the damage ratios corresponding to each of the segmented images to obtain the total damage ratio; Calculating the ratio of the total damage ratio to the number of the segmented images to obtain the average damage ratio; Performing a variance calculation on the damage ratios corresponding to each of the segmented images, the number of the segmented images, and the average damage ratio to obtain the damage ratio variance; Performing a summation calculation on the average damage ratio and the damage ratio variance to obtain the overall failure risk level of the conveying device.

9. The fault detection method according to claim 1, wherein The belt image includes the temperature image of the belt. The step of obtaining the belt image of the conveying device and identifying each damaged area in the belt image includes: Obtaining the temperature image of the belt in the conveying device; Inputting the temperature image into the trained damaged area recognition model to identify the abnormal temperature areas; determining the abnormal temperature list of the abnormal temperature areas based on the temperature values of each of the abnormal temperature areas; dividing the abnormal temperature list into multiple segmented temperature lists based on a preset temperature segmentation interval; determining the damaged temperature list based on a preset temperature determination threshold, a preset temperature determination parameter, and each of the segmented temperature lists; determining each damaged area in the temperature image based on the damaged temperatures in the damaged temperature list; wherein, the damaged area recognition model is trained by a temperature image training data set marked with the belt damaged areas.

10. The fault detection method according to claim 1, wherein The step of obtaining the belt image of the conveying device and identifying each damaged area in the belt image includes: Obtaining the belt image of the conveying device; Inputting the belt image into the trained damaged area recognition model to identify each damaged area in the belt image; wherein, the damaged area recognition model is trained by a belt image training data set marked with the damaged areas.

Images