Intelligent analysis method and system for equipment wear state fused with dynamic image sensing

Abstract

This invention relates to the field of computer vision technology, and discloses an intelligent analysis method and system for equipment wear status integrating dynamic image sensing. The method includes: performing frame-by-frame grayscale processing on a dynamic image sequence acquired during equipment operation to obtain a grayscale image set; extracting images of abrasive particle regions generated during equipment operation from the grayscale image set; identifying the abrasive particle concentration, volume ratio, and distortion degree of the abrasive particles produced by the equipment during operation based on the abrasive particle region images; analyzing the target core features of the abrasive particles based on the abrasive particle concentration; identifying the wear characteristics of the equipment based on the target core features; calculating the wear deviation degree of the abrasive particles based on the volume ratio and the distortion degree; and identifying the wear status of the equipment based on the wear deviation degree and the wear characteristics. This invention can improve the accuracy of equipment wear status analysis under different operating conditions.

Description

Technical Field

[0001] This invention relates to a method and system for intelligent analysis of device wear status by integrating dynamic image sensing, belonging to the field of computer vision technology. Background Technology

[0002] Equipment naturally wears down during operation. Analyzing the wear condition of equipment to assess its health is a common technique, used in the health monitoring of critical equipment such as CNC machine tools, wind power generation, and rail transportation.

[0003] Currently, equipment wear condition analysis often utilizes visual sensors to collect visual information such as surface texture, microscopic deformation, and vibration modes of equipment components, and then relies on historical experience to determine the specific wear condition. However, this method cannot adjust the analysis dimensions according to the stage of equipment wear evolution, resulting in a lack of accuracy under different operating conditions. Therefore, this solution utilizes wear particles generated during equipment wear that reflect the stage of equipment wear evolution to analyze the equipment wear condition. Summary of the Invention

[0004] This invention provides a method and system for intelligent analysis of equipment wear status by integrating dynamic image sensing, the main purpose of which is to improve the accuracy of equipment wear status analysis under different operating conditions.

[0005] To achieve the above objectives, the present invention provides an intelligent analysis method for device wear status fusion dynamic image sensing, comprising: The dynamic image sequence acquired during device operation is processed frame by frame into grayscale to obtain a grayscale image set, and the image of the abrasive region generated during device operation is extracted from the grayscale image set. Based on the abrasive region image, the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the device during operation are identified; Based on the abrasive concentration, the target core characteristics of the abrasive are analyzed, and based on the target core characteristics, the wear characteristics of the equipment are identified; Based on the body ratio and the deformity, the wear deviation of the abrasive grains is calculated, and based on the wear deviation and the wear characteristics, the wear state of the equipment is identified.

[0006] Optionally, based on the wear deviation and the wear characteristics, the wear state of the device is identified, including: Based on the wear characteristics, a preliminary analysis of the wear state of the equipment is conducted to obtain a preliminary classification set; Calculate the classification confidence of each category in the preliminary classification set; The confidence level of the classification is corrected using the wear deviation to obtain the corrected confidence level; Based on the corrected confidence level, the preliminary classification result is calibrated to obtain the target classification state, which is then used as the final wear state of the device.

[0007] Optionally, the step of using the wear deviation to correct the classification confidence level to obtain a corrected confidence level includes: The wear deviation is segmented and mapped to obtain a segmented correction factor; The piecewise correction factor is smoothed and optimized to obtain the optimized correction factor; The optimized correction factor is weighted and fused with the classification confidence to obtain the corrected confidence.

[0008] Optionally, based on the abrasive grain concentration, the target core characteristics of the abrasive grains are analyzed, including: The historical abrasive particle concentration values ​​of the device under different wear states in the historical operating data are queried, so as to construct the reference pattern of the abrasive particles based on the historical abrasive particle concentration values; The reference pattern and the abrasive concentration are subjected to gray target transformation to obtain the reference target center and the target center to be tested. Calculate the grey relational coefficients between the reference target center and the target center to be tested; Based on the gray relational coefficient, the target core characteristics of the abrasive grains are analyzed.

[0009] Optionally, the analysis of the target core characteristics of the abrasive grains based on the grey relational coefficient includes: Determine the influence coefficient of each interval on the wear state of the equipment based on the size range of the abrasive particles; Multiply each interval influence coefficient by the corresponding interval coefficient in the grey relational coefficient to obtain the weighted relational coefficient; The weighted correlation coefficients are mapped to the target center level of the abrasive particles; Based on the target center level, the target center characteristics of the abrasive grains are determined.

[0010] Optionally, based on the abrasive region image, the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the device during operation are identified, including: The total number of pixels identified as abrasive particles in the abrasive particle region image is counted. The abrasive particle concentration is obtained by calculating the ratio of the total number of pixels to the total number of pixels in the abrasive particle region image. Construct the bounding rectangle of the abrasive region image, and use the bounding rectangle to identify the volume ratio of the abrasive particles; Identify the abrasive grain perimeter and abrasive grain area to analyze the abrasive grain deformity based on the abrasive grain perimeter and abrasive grain area.

[0011] Optionally, extracting images of the abrasive region generated during device operation from the grayscale image set includes: Gaussian blurring is applied to each frame of the grayscale image set to obtain a noise-suppressed image; The Canny edge detection operator is used to detect the image edges of the noise-suppressed image to obtain the contour information of the abrasive particles; The contour information is binarized and segmented to obtain the abrasive grain region; Morphological processing is performed on the abrasive region to obtain an image of the abrasive region.

[0012] Optionally, based on the bullseye feature, the wear characteristics of the device are identified, including: Based on the bullseye features, the wear level of the device is constructed; Based on the wear level, match the wear keywords of the device; Based on the wear keywords, the wear characteristics of the device are identified.

[0013] Optionally, the wear deviation of the abrasive grains is calculated based on the body ratio and the deformity, including: The body proportion and the degree of deformity are weighted and fused to obtain a preliminary deviation. Query the equipment type and operating conditions of the device corresponding to the abrasive grains; Based on the equipment type and the operating conditions, the initial deviation is adjusted by a linear coefficient to obtain the wear deviation.

[0014] To address the aforementioned problems, this invention also provides an intelligent analysis system for device wear status integrating dynamic image sensing, the system comprising: The abrasive region recognition module is used to perform frame-by-frame grayscale processing on the dynamic image sequence acquired during equipment operation to obtain a grayscale image set, and extract the abrasive region image generated during equipment operation from the grayscale image set; The abrasive feature analysis module is used to identify the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the equipment during operation, based on the abrasive region image. The wear feature identification module is used to analyze the target feature of the abrasive particles based on the abrasive particle concentration, and to identify the wear features of the equipment based on the target feature; The wear condition analysis module is used to calculate the wear deviation of the abrasive particles based on the body ratio and the deformity, and to identify the wear condition of the equipment based on the wear deviation and the wear characteristics.

[0015] Compared to the problems described in the background technology, this invention first performs frame-by-frame grayscale processing on the dynamic image sequence of the device during operation and extracts the abrasive region image. This ensures the effectiveness of the analysis from the data source, because grayscale processing can effectively eliminate color interference in the original image, greatly simplify the data dimensions, reduce computational complexity, and avoid the obscuring of core features by multi-channel color information. The accurate extraction of the abrasive region is achieved by separating the abrasive target from the background and irrelevant structures, eliminating interference from redundant information such as the device body and operating environment, and firmly locking the focus of analysis on the abrasives directly related to wear. Then, based on the abrasive region image, this invention identifies abrasive concentration, body ratio, and deformity, transforming the visual morphological features of abrasives into three calculable and comparable quantitative parameters. Among them, abrasive concentration reflects the overall intensity of wear, body ratio characterizes the overall shape characteristics of abrasives, and deformity measures the irregularity of abrasive edge. The three together constitute a complete quantitative description system of abrasive morphology, upgrading wear analysis from "intuitive judgment" to "data-driven". Next, this invention constructs a benchmark model and gray target transformation, using target features to quantitatively compare the current abrasive particle concentration with the equipment's historical normal and typical wear states, accurately capturing the degree and direction of concentration deviation from the standard. Furthermore, through target level mapping and wear keyword matching, the specific type of wear is identified. Finally, this invention combines body ratio and deformity degree to calculate wear deviation and integrates wear features to identify the equipment's wear state, achieving a comprehensive judgment through multi-dimensional cross-validation, ensuring the reliability and practicality of the results. Therefore, this invention can improve the accuracy of equipment wear state analysis under different operating conditions. Attached Figure Description

[0016] Figure 1 A flowchart illustrating an intelligent analysis method for device wear status fusion with dynamic image sensing, provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of a module for implementing the intelligent analysis method for device wear status fusion dynamic image sensing according to an embodiment of the present invention; Figure 3 A schematic diagram of a computer device for an intelligent analysis method of device wear status fusion dynamic image sensing provided in an embodiment of the present invention; The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0018] This application provides a method for intelligent analysis of device wear status by integrating dynamic image sensing. The executing entity of this method includes, but is not limited to, at least one electronic device configured to execute the method provided in this application, such as a server or a terminal. In other words, the method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0019] Reference Figure 1 The diagram shown is a flowchart illustrating an intelligent analysis method for device wear status fusion based on dynamic image sensing, according to an embodiment of the present invention. In this embodiment, the intelligent analysis method for device wear status fusion based on dynamic image sensing includes: S1. Perform frame-by-frame grayscale processing on the dynamic image sequence collected during device operation to obtain a grayscale image set, and extract the image of the abrasive region generated during device operation from the grayscale image set.

[0020] The embodiments of the present invention perform frame-by-frame grayscale processing on the dynamic image sequence acquired during device operation to obtain a grayscale image set, which can eliminate color interference, thereby simplifying data dimensions and reducing computational complexity.

[0021] The equipment refers to a mechanical device that generates wear particles during operation, such as key moving parts like gears and bearings in metallurgical equipment, mining machinery, and generator sets. The dynamic image sequence refers to a set of images of continuous movement of the equipment collected by a dynamic sensor. The grayscale image set refers to a collection of images obtained by converting the dynamic image sequence to grayscale frame by frame.

[0022] Optionally, the grayscale image set can be obtained by performing frame-by-frame grayscale processing on the dynamic image sequence acquired during device operation using the RGB channel method.

[0023] Furthermore, in this embodiment of the invention, by extracting the abrasive region image generated during the operation of the device from the grayscale image set, the abrasive target region directly related to the wear of the device can be separated from the grayscale image set, eliminating interference from the background and other irrelevant structures.

[0024] The abrasive region image refers to a region image segmented from a grayscale image set that contains only abrasive targets.

[0025] As an embodiment of the present invention, extracting the image of the abrasive region generated during the operation of the device from the grayscale image set includes: Gaussian blurring is applied to each frame of the grayscale image set to obtain a noise-suppressed image; The Canny edge detection operator is used to detect the image edges of the noise-suppressed image to obtain the contour information of the abrasive particles; The contour information is binarized and segmented to obtain the abrasive grain region; Morphological processing is performed on the abrasive region to obtain an image of the abrasive region.

[0026] The noise-suppressed image refers to a grayscale image after Gaussian blurring, and the contour information refers to the geometric boundary formed by a series of connected pixels extracted from the noise-suppressed image by the Canny edge detection operator.

[0027] In detail, the noise-suppressed image can be obtained by convolving each frame of the grayscale image with a Gaussian kernel of a specific standard deviation; using the Canny edge detection operator, gradient calculation and non-maximum suppression are performed on the noise-suppressed image, and then the edge pixels are connected by a double thresholding method to detect complete image edges, thereby obtaining the contour information of the abrasive grains; the contour information is binarized and segmented to obtain the abrasive grain region, which can be processed using the OTSU adaptive thresholding algorithm; when performing morphological processing on the segmented abrasive grain region, dilation operation can be used to fill the tiny holes in the abrasive grain region, and then erosion operation can be used to remove the small noise at the edge of the region, repairing the incomplete parts of the abrasive grain contour, and finally obtaining the abrasive grain region image.

[0028] S2. Based on the abrasive region image, identify the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the device during operation.

[0029] This invention, through the identification of abrasive particle concentration, volume ratio, and deformity of abrasive particles produced by the equipment during operation based on the abrasive particle region image, can transform the visual morphology of abrasive particles into three quantifiable parameters: concentration, volume ratio, and deformity. This provides core data support for the analysis of equipment wear status.

[0030] The abrasive particle concentration refers to the ratio of the total area occupied by all wear particles in the abrasive particle region image to the total area of ​​the observable region in the image. The volume ratio refers to the ratio of the length of the long side to the short side of the minimum bounding rectangle of a single wear particle, used to describe whether the overall shape of the particle is slender or approximately circular. The deformity degree refers to the ratio of the perimeter of the edge of a single wear particle to its area, used to quantify the irregularity and complexity of the particle edge.

[0031] As an embodiment of the present invention, based on the abrasive region image, identifying the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the device during operation includes: The total number of pixels identified as abrasive particles in the abrasive particle region image is counted. The abrasive particle concentration is obtained by calculating the ratio of the total number of pixels to the total number of pixels in the abrasive particle region image. Construct the bounding rectangle of the abrasive region image, and use the bounding rectangle to identify the volume ratio of the abrasive particles; Identify the abrasive grain perimeter and abrasive grain area to analyze the abrasive grain deformity based on the abrasive grain perimeter and abrasive grain area.

[0032] In detail, the connected component labeling algorithm can be used to scan the abrasive region image, identify and count all interconnected abrasive pixel regions in the image, and determine the total number of pixels based on the statistical results. After obtaining the total number of abrasive pixels, it is divided by the total number of pixels in the abrasive region image to directly obtain the abrasive concentration value, i.e., the abrasive area ratio. First, the leftmost, rightmost, topmost, and bottommost boundary pixel coordinates of each independent abrasive region are located, and the smallest bounding rectangle that can completely enclose the abrasive is determined accordingly. Then, the length of the long side and the length of the short side of the rectangle are measured, and the ratio of the two is calculated as the abrasive volume ratio. The Suzuki contour tracking algorithm is used to extract the boundary pixel sequence of each abrasive region, and the abrasive perimeter is obtained by accumulating the Euclidean distance between adjacent boundary pixels. At the same time, the total number of pixels in the abrasive region is counted as the abrasive area.

[0033] Furthermore, as another embodiment of the present invention, the distortion degree can be calculated using the following formula: J = (L 2 / S 2 ); Where J represents the degree of distortion, L represents the perimeter, and S represents the area.

[0034] S3. Based on the abrasive concentration, analyze the target characteristics of the abrasive particles, and based on the target characteristics, identify the wear characteristics of the equipment.

[0035] This invention, through analysis of the target characteristics of the abrasive particles based on the abrasive particle concentration, can determine the difference between the abrasive particle concentration of the equipment and the standard state, thereby identifying the degree of deviation between the current wear characteristics of the equipment and the normal level.

[0036] The bullseye feature refers to a quantitative indicator used to characterize the degree of closeness between the current wear state of the equipment and the preset standard state.

[0037] As an embodiment of the present invention, the target core characteristics of the abrasive grains are analyzed based on the abrasive grain concentration, including: The historical abrasive particle concentration values ​​of the device under different wear states in the historical operating data are queried, so as to construct the reference pattern of the abrasive particles based on the historical abrasive particle concentration values; The reference pattern and the abrasive concentration are subjected to gray target transformation to obtain the reference target center and the target center to be tested. Calculate the grey relational coefficients between the reference target center and the target center to be tested; Based on the gray relational coefficient, the target core characteristics of the abrasive grains are analyzed.

[0038] The reference pattern refers to a set of standard abrasive particle concentration values ​​extracted from the historical operating data of the equipment, which can represent different typical wear states; the grey relational coefficient refers to the coefficient used to quantify the similarity between the current abrasive particle concentration value and each reference pattern sequence on the corresponding elements.

[0039] In detail, representative historical abrasive concentration data under different wear states can be extracted from the equipment's historical operation database. The concentration values ​​corresponding to the three typical states of normal wear, slight wear, and severe wear in the historical abrasive concentration data can be identified, and a reference pattern can be constructed based on this concentration value. When performing gray target transformation, the currently measured abrasive concentration and the reference pattern sequence can be divided into the same intervals to obtain a test pattern corresponding to the reference pattern. Then, the standard concentration value and the current abrasive concentration value in the same interval of the reference pattern and the test pattern can be calculated, and the minimum and maximum values ​​of the two can be determined respectively. Then, the standard concentration value and the current abrasive concentration value in each interval can be transformed by dividing the minimum value by the maximum value. Finally, the transformed values ​​of all size intervals obtained by the above calculation are integrated in the original size interval arrangement order to obtain the reference target center and the test target center.

[0040] Furthermore, as another embodiment of the present invention, the formula for calculating the grey relational coefficient is as follows: ; in, Represents the gray target correlation coefficient. The standard bullseye is at the 1st The value of each indicator, Indicates the target center at the th ... The value of each indicator, This indicates the standard bullseye and the target bullseye at the 1st... The absolute difference of each indicator This represents the resolution coefficient.

[0041] Preferably, the step of analyzing the target core characteristics of the abrasive grains based on the grey relational coefficient includes: Determine the influence coefficient of each interval on the wear state of the equipment based on the size range of the abrasive particles; Multiply each interval influence coefficient by the corresponding interval coefficient in the grey relational coefficient to obtain the weighted relational coefficient; The weighted correlation coefficients are mapped to the target center level of the abrasive particles; Based on the target center level, the target center characteristics of the abrasive grains are determined.

[0042] The influence coefficients of each interval refer to a set of weight values ​​determined by historical data analysis and statistics based on the degree of influence of abrasive particles of different size ranges on the wear state of the equipment. The target level refers to a classification index used to characterize the degree of closeness between the current wear state and the ideal state.

[0043] In detail, by combining equipment wear mechanisms and historical operating data, the frequency of equipment failures caused by abrasive particles in different size ranges and the contribution ratio of wear severity can be statistically analyzed. A corresponding influence coefficient can be assigned to each abrasive particle size range. Specifically, the Pearson correlation coefficient method can be used to calculate the statistical correlation between the change in abrasive particle concentration and the degree of wear in each size range, and the magnitude of the coefficient is positively correlated with the degree of influence of abrasive particles in that range on equipment wear. Based on a preset target intensity grading threshold, the comprehensive target intensity is mapped to a specific target intensity level. For example, (0.8, 1.0) is set as the normal wear level, (0.6, 0.8) as the slight wear level, (0.4, 0.6) as the severe wear level, and <0.4 as the catastrophic wear level. Based on the mapped target intensity level, combined with the typical wear characteristics corresponding to that level, the target intensity characteristic description of the abrasive particles is determined. For example, the target intensity characteristic corresponding to the normal wear level can be described as "uniform abrasive particle distribution, mainly small abrasive particles," while the target intensity characteristic corresponding to the severe wear level can be described as "a significant increase in the proportion of large-sized abrasive particles, with concentrated distribution."

[0044] Furthermore, in this embodiment of the invention, by identifying the wear characteristics of the device based on the bullseye feature, the specific pattern and nature of the device wear can be determined based on the degree of deviation reflected by the bullseye feature.

[0045] The wear characteristics refer to the basic pattern or type of equipment wear, such as "normal wear", "slight wear" or "severe wear".

[0046] As an embodiment of the present invention, identifying the wear characteristics of the device based on the target feature includes: Based on the bullseye features, the wear level of the device is constructed; Based on the wear level, match the wear keywords of the device; Based on the wear keywords, the wear characteristics of the device are identified.

[0047] The wear level refers to a discrete grading index used to characterize the severity of equipment wear, which is divided according to the preset threshold range of the target feature value. For example, level 1 (normal wear), level 2 (slight wear), and level 3 (severe wear). The wear keywords refer to the typical damage morphology features of the equipment at different levels, such as "smooth surface", "slight scratches", and "deep grooves".

[0048] In detail, the target level can be directly mapped to the wear level. For example, normal wear level is mapped to level 1 wear, slight wear level is mapped to level 2 wear, and severe wear level is mapped to level 3 wear. Based on the determined wear level, the historical wear feature terminology is accessed, and the corresponding feature description terms are extracted using keyword mapping technology. For example, the wear keywords corresponding to level 1 wear are uniform size and dispersed distribution, and the corresponding wear features are uniform wear and trace wear. The wear keywords corresponding to level 2 wear are medium size and a small number of large particles, and the corresponding wear features are adhesive wear and slight scratches. The wear keywords corresponding to level 3 wear are large-sized particles and clustered distribution, and the corresponding wear features are corrosive wear and surface peeling.

[0049] S4. Calculate the wear deviation of the abrasive particles based on the body ratio and the deformity, and identify the wear state of the equipment based on the wear deviation and the wear characteristics.

[0050] The embodiments of the present invention calculate the wear deviation of the abrasive grains based on the body ratio and the deformity, which can quantify the overall abnormality of the abrasive grain shape and provide an intuitive and comprehensive morphological basis for judging the severity of the wear state.

[0051] The wear deviation refers to the degree of abnormality in the shape of wear particles as a whole.

[0052] As an embodiment of the present invention, the wear deviation of the abrasive grains is calculated based on the body ratio and the deformity, including: The body proportion and the degree of deformity are weighted and fused to obtain a preliminary deviation. Query the equipment type and operating conditions of the device corresponding to the abrasive grains; Based on the equipment type and the operating conditions, the initial deviation is adjusted by a linear coefficient to obtain the wear deviation.

[0053] The "equipment type" refers to the specific category of the industrial equipment that generates the abrasive particles. This classification is based on the equipment's core functions, structural features, and working principles, and specifically includes rolling bearings, gear transmission devices, hydraulic systems, and reciprocating compressors, etc. The "operating conditions" refer to the equipment's real-time operating environment and working status parameters, including operating load, operating temperature, lubrication status, operating time, and external environmental factors, etc. In detail, the body proportions and deformity degree can first be normalized to eliminate dimensional differences, and then a preliminary deviation can be calculated using a linear weighted formula, which can be used as follows: D = w1 * T + w2 * J; Where D represents the initial deviation, T represents the body proportion, J represents the deformity, w1 represents the weighting coefficient corresponding to T, and w2 represents the weighting coefficient corresponding to J. Furthermore, the parameter database of the equipment management system can be queried to retrieve the basic information of the equipment currently generating abrasive particles, clarifying the equipment type and core technical parameters. At the same time, real-time operating data of the equipment can be collected to extract key operating conditions, including operating load, operating temperature, lubrication status, running time, and working medium characteristics. Based on historical operating data, a corresponding relationship database of different equipment types, different operating conditions, and deviation correction coefficients can be established. The setting of the correction coefficient needs to be combined with the allowable deviation range of abrasive particle morphology parameters under specific operating conditions (e.g., the correction coefficient under heavy load conditions is greater than that under light load conditions to enhance deviation sensitivity). Then, according to the retrieved equipment type and operating conditions, a suitable linear adjustment coefficient is matched from the corresponding relationship database. The initial deviation is multiplied by the adjustment coefficient to obtain the final wear deviation.

[0054] Furthermore, in this embodiment of the invention, the wear state of the device can be identified by integrating morphological anomaly indicators and macroscopic distribution characteristics based on the wear deviation degree and the wear characteristics, and a comprehensive determination of the wear state can be achieved through multi-dimensional feature cross-validation.

[0055] The wear condition refers to the typical state category of the equipment due to the degree of wear caused by long-term operation, such as "normal wear", "slight wear" or "severe wear".

[0056] As an embodiment of the present invention, identifying the wear state of the equipment based on the wear deviation and the wear characteristics includes: Based on the wear characteristics, a preliminary analysis of the wear state of the equipment is conducted to obtain a preliminary classification set; Calculate the classification confidence of each category in the preliminary classification set; The confidence level of the classification is corrected using the wear deviation to obtain the corrected confidence level; Based on the corrected confidence level, the preliminary classification result is calibrated to obtain the target classification state, which is then used as the final wear state of the device.

[0057] The preliminary classification set refers to the set of possible wear states of equipment obtained based on wear characteristic analysis. The classification confidence refers to the quantitative value of the reliability of the judgment results of each wear state in the preliminary classification set. The corrected confidence refers to the optimized estimate obtained after correcting the original classification confidence using wear deviation.

[0058] In detail, support vector machines (SVMs) can be used to perform pattern recognition on wear features, transforming the input feature descriptions into specific wear state classification results. For example, normal wear is characterized by abrasive particle concentration within the baseline range, a volume ratio close to 1:1, and low distortion, corresponding to a stable equipment operation phase. Slight wear is characterized by a slightly increased abrasive particle concentration, the appearance of elongated particles with a volume ratio of 1.5-2.0, and an increase in distortion. Severe wear is characterized by a significantly increased abrasive particle concentration, an increase in flaky particles with a volume ratio exceeding 3.0, and a marked increase in distortion. When calculating classification confidence, the softmax function can be used to convert the SVM output into standardized confidence values. When performing final calibration of the initial classification results based on the corrected confidence, low-confidence classification results can be re-evaluated or labeled by setting a confidence threshold. By comparing the corrected confidence of each classification result, the state with the highest confidence is selected as the final judgment result.

[0059] Preferably, the step of using the wear deviation to correct the classification confidence level to obtain a corrected confidence level includes: The wear deviation is segmented and mapped to obtain a segmented correction factor; The piecewise correction factor is smoothed and optimized to obtain the optimized correction factor; The optimized correction factor is weighted and fused with the classification confidence to obtain the corrected confidence.

[0060] The segmentation correction factor refers to a parameter that determines the degree of abrasive grain morphology abnormality in relation to the classification confidence level.

[0061] In detail, segmentation rules can be set according to the range of deviation using a piecewise function: when the deviation is within the normal threshold range, it is mapped to 1.0 (no correction required); when it is between the upper limit of the normal threshold and the lower limit of the abnormal threshold, it is mapped to a value between 0.5 and 1.0 according to a linear decreasing rule; when it exceeds the lower limit of the abnormal threshold, it is mapped to a value between 0.1 and 0.5 (the value decreases as the deviation increases). When smoothing and optimizing the piecewise correction factor, a sigmoid function, such as the sigmoid function, can be used to smooth and optimize the initially calculated correction factor.

[0062] Furthermore, as another embodiment of the present invention, the formula for calculating the corrected confidence level is as follows: ; in, This indicates the adjusted confidence level. Indicates classification confidence level. Represents the integer weighting coefficient. This represents the optimization correction factor.

[0063] It should be further noted that the weighting coefficients need to be dynamically determined based on the equipment type and operating conditions.

[0064] like Figure 2 The diagram shown is a functional module diagram of the intelligent analysis system for equipment wear status that integrates dynamic image sensing according to the present invention.

[0065] The intelligent device wear status analysis system 200 integrating dynamic image sensing described in this invention can be installed in electronic devices. Depending on the functions implemented, the intelligent device wear status analysis system integrating dynamic image sensing can control the wear particle region identification module 201, the wear particle feature analysis module 202, the wear feature identification module 203, and the wear status analysis module 204. The module described in this invention can also be called a unit, referring to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0066] In this embodiment of the invention, the functions of each module / unit are as follows: The abrasive region identification module 201 is used to perform frame-by-frame grayscale processing on the dynamic image sequence collected during equipment operation to obtain a grayscale image set, and extract the abrasive region image generated during equipment operation from the grayscale image set; The abrasive feature analysis module 202 is used to identify the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the equipment during operation based on the abrasive region image. The wear feature identification module 203 is used to analyze the target feature of the abrasive particles based on the abrasive particle concentration, and to identify the wear features of the equipment based on the target feature; The wear state analysis module 204 is used to calculate the wear deviation of the abrasive particles based on the body ratio and the deformity, and to identify the wear state of the equipment based on the wear deviation and the wear characteristics.

[0067] In detail, the modules in the intelligent device wear status analysis system 200 that integrates dynamic image sensing described in this embodiment of the invention employ the same methods as described above during use. Figure 1 The method used here is the same as the intelligent analysis method for device wear status that integrates dynamic image sensing, and can produce the same technical effect, so it will not be elaborated here.

[0068] In one embodiment, a computer device is provided, which may be a server or a client, and its internal structure diagram may be as follows: Figure 3As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external clients via a network connection. When the computer program is executed by the processor, it implements the functions or steps of a device wear status intelligent analysis method integrating dynamic image sensing on the server or client side.

[0069] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: The dynamic image sequence acquired during device operation is processed frame by frame into grayscale to obtain a grayscale image set, and the image of the abrasive region generated during device operation is extracted from the grayscale image set. Based on the abrasive region image, the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the device during operation are identified; Based on the abrasive concentration, the target core characteristics of the abrasive are analyzed, and based on the target core characteristics, the wear characteristics of the equipment are identified; Based on the body ratio and the deformity, the wear deviation of the abrasive grains is calculated, and based on the wear deviation and the wear characteristics, the wear state of the equipment is identified.

[0070] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: The dynamic image sequence acquired during device operation is processed frame by frame into grayscale to obtain a grayscale image set, and the image of the abrasive region generated during device operation is extracted from the grayscale image set. Based on the abrasive region image, the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the device during operation are identified; Based on the abrasive concentration, the target core characteristics of the abrasive are analyzed, and based on the target core characteristics, the wear characteristics of the equipment are identified; Based on the body ratio and the deformity, the wear deviation of the abrasive grains is calculated, and based on the wear deviation and the wear characteristics, the wear state of the equipment is identified.

[0071] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0072] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0073] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0074] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0075] Finally, it should be noted that in the above embodiments, each embodiment can be combined with each other or independent. Deleting any one of them will not affect the technical implementation of other embodiments. The above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for intelligent analysis of equipment wear status integrating dynamic image sensing, characterized in that, The method includes: The dynamic image sequence acquired during device operation is processed frame by frame into grayscale to obtain a grayscale image set, and the image of the abrasive region generated during device operation is extracted from the grayscale image set. Based on the abrasive region image, the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the device during operation are identified; Based on the abrasive concentration, the target core characteristics of the abrasive are analyzed, and based on the target core characteristics, the wear characteristics of the equipment are identified; Based on the body ratio and the deformity, the wear deviation of the abrasive grains is calculated, and based on the wear deviation and the wear characteristics, the wear state of the equipment is identified.

2. The intelligent analysis method for device wear status fusion with dynamic image sensing as described in claim 1, characterized in that, Based on the wear deviation and the wear characteristics, the wear state of the equipment is identified, including: Based on the wear characteristics, a preliminary analysis of the wear state of the equipment is conducted to obtain a preliminary classification set; Calculate the classification confidence of each category in the preliminary classification set; The confidence level of the classification is corrected using the wear deviation to obtain the corrected confidence level; Based on the corrected confidence level, the preliminary classification result is calibrated to obtain the target classification state, which is then used as the final wear state of the device.

3. The intelligent analysis method for device wear status fusion with dynamic image sensing as described in claim 2, characterized in that, Using the wear deviation, the classification confidence level is corrected to obtain the corrected confidence level, which includes: The wear deviation is segmented and mapped to obtain a segmented correction factor; The piecewise correction factor is smoothed and optimized to obtain the optimized correction factor; The optimized correction factor is weighted and fused with the classification confidence to obtain the corrected confidence.

4. The intelligent analysis method for device wear status fusion with dynamic image sensing as described in claim 1, characterized in that, Based on the abrasive grain concentration, the target core characteristics of the abrasive grains are analyzed, including: The historical abrasive particle concentration values ​​of the device under different wear states in the historical operating data are queried, so as to construct the reference pattern of the abrasive particles based on the historical abrasive particle concentration values; The reference pattern and the abrasive concentration are subjected to gray target transformation to obtain the reference target center and the target center to be tested. Calculate the grey relational coefficients between the reference target center and the target center to be tested; Based on the gray relational coefficient, the target core characteristics of the abrasive grains are analyzed.

5. The intelligent analysis method for device wear status fusion with dynamic image sensing as described in claim 4, characterized in that, Based on the gray relational coefficient, the target core characteristics of the abrasive grains are analyzed, including: Determine the influence coefficient of each interval on the wear state of the equipment based on the size range of the abrasive particles; Multiply each interval influence coefficient by the corresponding interval coefficient in the grey relational coefficient to obtain the weighted relational coefficient; The weighted correlation coefficients are mapped to the target center level of the abrasive particles; Based on the target center level, the target center characteristics of the abrasive grains are determined.

6. The intelligent analysis method for device wear status fusion with dynamic image sensing as described in claim 1, characterized in that, Based on the abrasive region image, the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the device during operation are identified, including: The total number of pixels identified as abrasive particles in the abrasive particle region image is counted. The abrasive particle concentration is obtained by calculating the ratio of the total number of pixels to the total number of pixels in the abrasive particle region image. Construct the bounding rectangle of the abrasive region image, and use the bounding rectangle to identify the volume ratio of the abrasive particles; Identify the abrasive grain perimeter and abrasive grain area to analyze the abrasive grain deformity based on the abrasive grain perimeter and abrasive grain area.

7. The intelligent analysis method for device wear status fusion with dynamic image sensing as described in claim 1, characterized in that, Extracting the image of the abrasive region generated during the operation of the device from the grayscale image set includes: Gaussian blurring is applied to each frame of the grayscale image set to obtain a noise-suppressed image; The Canny edge detection operator is used to detect the image edges of the noise-suppressed image to obtain the contour information of the abrasive particles; The contour information is binarized and segmented to obtain the abrasive grain region; Morphological processing is performed on the abrasive region to obtain an image of the abrasive region.

8. The intelligent analysis method for device wear status fusion with dynamic image sensing as described in claim 1, characterized in that, Based on the bullseye feature, the wear characteristics of the device are identified, including: Based on the bullseye features, the wear level of the device is constructed; Based on the wear level, match the wear keywords of the device; Based on the wear keywords, the wear characteristics of the device are identified.

9. The intelligent analysis method for device wear status fusion with dynamic image sensing as described in claim 1, characterized in that, Based on the body ratio and the deformity, the wear deviation of the abrasive grains is calculated, including: The body proportion and the degree of deformity are weighted and fused to obtain a preliminary deviation. Query the equipment type and operating conditions of the device corresponding to the abrasive grains; Based on the equipment type and the operating conditions, the initial deviation is adjusted by a linear coefficient to obtain the wear deviation.

10. An intelligent analysis system for equipment wear status integrating dynamic image sensing, characterized in that, The system includes: The abrasive region recognition module is used to perform frame-by-frame grayscale processing on the dynamic image sequence acquired during equipment operation to obtain a grayscale image set, and extract the abrasive region image generated during equipment operation from the grayscale image set; The abrasive feature analysis module is used to identify the abrasive concentration, volume ratio, and deformity of the abrasive particles produced by the equipment during operation, based on the abrasive region image. The wear feature identification module is used to analyze the target feature of the abrasive particles based on the abrasive particle concentration, and to identify the wear features of the equipment based on the target feature; The wear condition analysis module is used to calculate the wear deviation of the abrasive particles based on the body ratio and the deformity, and to identify the wear condition of the equipment based on the wear deviation and the wear characteristics.

Images