Intelligent detection method for quality of spherical alumina

Abstract

The invention relates to the technical field of spherical aluminum oxide quality detection, in particular to an intelligent detection method for spherical aluminum oxide quality, which comprises the following steps: collecting microscopic images and chemical component purity of a spherical aluminum oxide sample; outputting the sphericity degree and defect proportion of the spherical alumina sample; outputting a quality grade evaluation result; carrying out fluidity test on the spherical aluminum oxide sample, sequentially carrying out acquisition, comparison and auditing on image data of the spherical aluminum oxide sample, and calculating the flow velocity of the spherical aluminum oxide sample; determining a detection accuracy adjustment mode according to the velocity variance of the flow velocities of the plurality of spherical alumina samples, including adjusting the comparison type of the image data, or adjusting the auditing interval number of the image data; wherein the shooting frequency for collecting the image data is determined according to the instantaneous speed variation of the collision of the spherical aluminum oxide sample. According to the invention, the accuracy of alumina detection is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of quality detection of spherical alumina, and particularly to an intelligent detection method for the quality of spherical alumina. Background Art

[0002] In the prior art, a high-resolution industrial camera is combined to take microscopic images, and an image processing algorithm is used to calculate the area perimeter ratio to determine the qualified spherical standard; by evaluating roughness and measuring specific surface area, it is applicable to high-performance application scenarios such as catalyst carriers.

[0003] Chinese Patent Publication No.: CN119580366A discloses an intelligent data analysis system and method based on alumina production, including: a parameter acquisition module, which acquires ball mill parameters, raw material parameters, and spherical alumina impurity thresholds; identifies the ball mill parameters and raw material parameters according to a threshold prediction model to determine standard thresholds; a sample acquisition module, which processes alumina raw materials through a ball mill to obtain spherical alumina; detects the purity of the spherical alumina, and samples the spherical alumina by random sampling to obtain multiple samples; observes the samples through an electron microscope to obtain a set of morphology images; an image segmentation module, which constructs a morphology image segmentation model to segment the set of morphology images to obtain a first image set and a second image set; the first image set is the microscopic morphology of spherical alumina, and the second image set is the microscopic morphology of impurities; a first recognition module, which constructs a first image recognition model to recognize the first image set to obtain a set of spherical alumina parameters; the set of spherical alumina parameters includes particle size data, sphericity data, and smoothness coefficient of spherical alumina particles; calculates a first coefficient according to the set of spherical alumina parameters; a second recognition module, which constructs a second image recognition model to recognize the second image set to obtain a set of impurity parameters; the set of impurity parameters includes impurity particle size, impurity sphericity, and impurity smoothness coefficient; calculates a second coefficient according to the set of impurity parameters; a comprehensive evaluation module, which calculates a quality coefficient through the first coefficient, the second coefficient, the spherical alumina impurity threshold, and the spherical alumina purity. It can be seen that the intelligent data analysis system and method based on alumina production have problems such as insufficient comprehensiveness of data acquisition due to inconsistent interference of the surrounding structure of the acquisition environment on image acquisition caused by unstable image acquisition environment, chaotic modeling and low efficiency due to redundant data participating in modeling or data participating in model update, or decreased modeling accuracy due to undetected abnormal data in the modeling stage. Summary of the Invention

[0004] To this end, the present invention provides an intelligent detection method for the quality of spherical alumina, which is used to overcome the problems in the prior art that due to the unstable image acquisition environment, the interference of the surrounding structures of the acquisition environment to image acquisition is inconsistent, resulting in insufficient comprehensiveness of data acquisition; due to the redundancy of the data participating in model building or the data participating in model update, the model building is chaotic and the efficiency is low; or due to the missed detection of abnormal data in the model building stage, the accuracy of model building decreases.

[0005] To achieve the above object, the present invention provides an intelligent detection method for the quality of spherical alumina, including: Collect the microscopic images and chemical composition purity of spherical alumina samples; Segment and compare the microscopic images to obtain the characteristic parameters of the spherical alumina samples, and output the sphericity and defect ratio of the spherical alumina samples. Among them, the characteristic parameters include the particle size and defect area of the spherical alumina samples; Respectively input the sphericity, the defect ratio and the chemical composition purity of the spherical alumina samples into a multi-modal fusion model, and output the quality grade evaluation result; Conduct a fluidity test on the spherical alumina samples, and sequentially collect, compare and audit the image data of the spherical alumina samples, and calculate the flow velocity of the spherical alumina samples; Determine the detection accuracy adjustment method according to the velocity variance of the flow velocities of several spherical alumina samples, including adjusting the comparison type of the image data, or adjusting the audit interval number of the image data according to the grade difference between the output grade of the model and the actual quality grade of the spherical alumina; Among them, the shooting frequency of collecting the image data is determined according to the instantaneous velocity change amount of the collision of the spherical alumina samples.

[0006] Further, the process of adjusting the comparison type of the image data includes: Under the same conditions, collect the flow velocities of spherical alumina samples of the same batch on an inclined slideway several times; Calculate the velocity variance according to several flow velocities; If the velocity variance is greater than or equal to a preset first variance, it is determined that the accuracy of model establishment does not meet the requirements; If the velocity variance is greater than or equal to a preset second variance, it is determined that the comprehensiveness of data acquisition does not meet the requirements, and the number of comparison types of the collected image data is increased; Among them, the preset first variance is less than the preset second variance; the number of comparison types of the collected image data is positively correlated with the velocity variance.

[0007] Further, the same conditions include that the volumes of the spherical alumina samples of the same batch for performing the fluidity test are the same, the inclination angles of the inclined chutes are the same, and the ambient temperatures of the fluidity test are the same.

[0008] Further, the process of adjusting the audit interval number of the image data includes: If the velocity variance is greater than or equal to the preset first variance and less than or equal to the preset second variance, it is preliminarily determined that the accuracy of the data participating in the modeling does not meet the requirements; Obtain the output grade of the model and the actual quality grade of the spherical alumina; According to the grade difference between the output grade of the model and the actual quality grade of the spherical alumina; If the grade difference is greater than the preset grade difference, it is secondarily determined that the accuracy of the data participating in the modeling does not meet the requirements, and the audit interval number of the data is increased; Wherein, the audit interval number of the data is positively correlated with the grade difference.

[0009] Further, the grade difference is the absolute value of the difference between the output grade of the model and the actual quality grade of the spherical alumina.

[0010] Further, the process of determining the shooting frequency of collecting the image data according to the change amount of the instantaneous speed of the collision of the spherical alumina samples includes: Respectively obtain the first displacement amount of the alumina that has collided within the unit fluidity test duration before the collision and the second displacement amount within the unit fluidity test duration after the collision; Calculate the change amount of the instantaneous speed according to the first displacement amount and the second displacement amount; If the change amount of the instantaneous speed is greater than the preset change amount, it is determined that the real-time performance of the image acquisition does not meet the requirements, and the shooting frequency of collecting the image data is increased.

[0011] Further, the change amount of the instantaneous speed is the difference between the instantaneous speed before the collision and the instantaneous speed after the collision.

[0012] Further, the instantaneous speed before the collision is the ratio of the first displacement amount to the unit fluidity test duration; the instantaneous speed after the collision is the ratio of the second displacement amount to the unit fluidity test duration.

[0013] Further, the process of determining the unit fluidity test duration includes: Obtain the passing moments of the marked particle flows of the spherical alumina samples at the ends of several inclined chutes; Calculate the standard deviation of the time intervals of several said passing moments; If the standard deviation of the time interval is greater than or equal to a preset first standard deviation and less than a preset second standard deviation, the unit liquidity test duration is set to a first-level duration; If the standard deviation of the time interval is greater than or equal to the preset second standard deviation and less than a preset third standard deviation, the unit liquidity test duration is set to a second-level duration; If the standard deviation of the time interval is greater than or equal to the preset third standard deviation, the unit liquidity test duration is set to a third-level duration.

[0014] Furthermore, the first-level duration is greater than the second-level duration, and the second-level duration is greater than the third-level duration.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows. The method of the present invention constructs an intelligent detection system for multimodal fusion; through cross-modal data fusion of a high-resolution industrial camera, a laser scatterometer, and X-ray fluorescence spectroscopy, synchronous analysis of the morphological characteristics, particle size distribution, and chemical composition of spherical alumina is achieved; by establishing an association between the dynamic parameters of the flow velocity and the image acquisition frequency, modeling is realized. Due to the elastic deformation after particle collision resulting in the loss of key deformation frames, which in turn leads to a decrease in the detection rate of surface defects, or due to a large number of repeated images being introduced when the fluidity is stable, increasing the computational burden and introducing noise, and the displacement measurement window being too long, resulting in the problem of being unable to capture the transient switching of particle stagnation-flow, and the problem of decreased modeling accuracy due to the missed detection of abnormal data during the modeling stage, by adjusting the shooting frequency in real time according to the instantaneous velocity change amount before and after the collision, the effectiveness of image acquisition is improved, by adjusting the number of audit intervals of the data, the misjudgment of a single index is reduced, and the modeling accuracy is improved, and by adjusting the comparison type of the image data, the comprehensiveness of data acquisition is improved.

[0016] Furthermore, the method of the present invention sets a first preset variance and a second preset variance. Due to the problem that surface microcracks in the dark field and grain boundary anomalies in polarized light are prone to be missed during only bright-field imaging, for example, the interference of the surrounding structure on image acquisition is inconsistent due to different acquisition angles, and the missing detection of features leads to an increase in the misjudgment rate of the model, which in turn causes the morphological and compositional data of the same particle to be unable to match and the multimodal fusion accuracy to decrease. By increasing the comparison type, the accuracy and comprehensiveness of image acquisition are improved.

[0017] Furthermore, the method of the present invention sets a preset level difference amount. For the micro-displacement at the moment of collision, such as elastic deformation that cannot be accurately captured, and the misdetection of components due to XRF drift, by increasing the number of audit intervals, the problem of decreased accuracy of the model output result caused by the misjudgment of a single index is reduced, and the accuracy of the model is improved.

[0018] Furthermore, in the method of the present invention, by setting a preset change amount, due to the elastic deformation after particle collision, key deformation frames are lost, resulting in a decrease in the detection rate of surface defects. Or, when the fluidity is stable, a large number of repeated images are introduced, increasing the computational burden and introducing noise, resulting in delays or errors in image acquisition, and further leading to an iterative increase in the error amount of the participating data in the initial stage of modeling. By increasing the shooting frequency, an increase in modeling accuracy is achieved.

[0019] Furthermore, the present invention determines the unit fluidity test duration, calculates the standard deviation of the particle flow time interval, and establishes a three-level duration determination method. Through adaptive timing control, the best sampling rate can still be maintained when the powder fluidity fluctuation is detected. By constructing a reverse duration adjustment mechanism, the sampling conditions are extended when the fluidity is poor to ensure data integrity and increase detection accuracy. When the fluidity is good, the duration is shortened to improve timeliness and reduce system power consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is the overall flowchart of the intelligent detection method for the quality of spherical alumina in the embodiment of the present invention; Figure 2 It is the contrast type adjustment flowchart of the intelligent detection method for the quality of spherical alumina in the embodiment of the present invention; Figure 3 It is the flowchart for adjusting the audit interval number of image data in the intelligent detection method for the quality of spherical alumina in the embodiment of the present invention; Figure 4 It is the flowchart for determining the shooting frequency of the industrial camera in the intelligent detection method for the quality of spherical alumina in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] In order to make the objectives and advantages of the present invention clearer, the present invention will be further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0022] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present invention and do not limit the protection scope of the present invention.

[0023] It should be noted that in the description of the present invention, the terms indicating directions or positional relationships such as "up", "down", "left", "right", "inside", "outside", etc. are based on the directions or positional relationships shown in the drawings. This is only for convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

[0024] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and defined, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, and it can be the communication inside two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0025] Please refer to Figure 1 , Figure 2 , Figure 3 and Figure 4 as shown, which are respectively the overall flowchart, the comparison type adjustment flowchart, the flowchart for adjusting the audit interval number of image data, and the flowchart for determining the shooting frequency of the industrial camera of the intelligent detection method for the quality of spherical alumina in the embodiments of the present invention. An intelligent detection method for the quality of spherical alumina in an embodiment of the present invention includes: Collect the microscopic images and chemical composition purity of spherical alumina samples; Segment and compare the microscopic images to obtain the characteristic parameters of the spherical alumina samples to output the sphericity and defect ratio of the spherical alumina samples, where the characteristic parameters include the particle size and defect area of the spherical alumina samples; Input the sphericity, the defect ratio, and the chemical composition purity of the spherical alumina samples into a multi-modal fusion model respectively, and output a quality grade evaluation result; Conduct a fluidity test on the spherical alumina samples, and sequentially collect, compare and audit the image data of the spherical alumina samples and calculate the flow velocity of the spherical alumina samples; Determine the detection accuracy adjustment method according to the velocity variance of the flow velocities of several spherical alumina samples, including adjusting the comparison type of the image data, or adjusting the audit interval number of the image data according to the grade difference between the output grade of the model and the actual quality grade of the spherical alumina; Among them, the shooting frequency for collecting the image data is determined according to the instantaneous velocity change amount of the collision of the spherical alumina samples.

[0026] Specifically, the spherical alumina samples are granular.

[0027] Specifically, a high-resolution industrial camera is used to collect the microscopic images of spherical alumina samples. Examples of the models of high-resolution industrial cameras include Photron FASTCAM Nova S12 and UC1400 industrial microscopic cameras.

[0028] Specifically, the chemical composition purity of the spherical alumina sample is the molar mass content of alumina in the spherical alumina sample, and the spherical alumina sample also contains silicon dioxide and iron oxide; the chemical composition purity of the spherical alumina sample is collected by an elemental analyzer.

[0029] Specifically, the microscopic image is segmented and compared to extract the characteristic parameters of the spherical alumina sample, and the sphericity, surface roughness and defect ratio of the spherical alumina sample are calculated according to the convolutional neural network model, including: Obtain the microscopic image of the spherical alumina sample, where a single particle occupies at least a 50×50 pixel area; The U-Net segmentation network segments the microscopic image and compares it with the images of spherical alumina samples in the historical database according to the comparison type to obtain the defect area; Calculate the sphericity of the segmented microscopic image according to the geometric method, and determine it as a qualified sphere when the sphericity is greater than or equal to 0.92; Use Mask R-CNN to determine the defect area of the segmented microscopic image, where the defect area includes areas where the crack aspect ratio is greater than or equal to 3, areas where the circularity of the hole is greater than or equal to 0.85, and areas where the edge gradient is greater than or equal to 15%; the defect ratio is the ratio of the total area of the defect area to the area of the spherical alumina sample in the microscopic image.

[0030] Specifically, the defect area includes the absence and depression areas of spherical alumina.

[0031] Specifically, the sphericity, the defect ratio and the chemical composition purity of the spherical alumina sample are input into the multi-modal fusion model, and the quality grade evaluation result is output, including: Perform data standardization processing on the sphericity, defect ratio and chemical composition purity of the spherical alumina sample; Calculate the weights of the sphericity, defect ratio and chemical composition purity after data standardization through a two-layer neural network, and use the weighted average method to calculate: Determine the quality grade according to the calculated result.

[0032] In implementation, the method of the present invention classifies the quality grade into grade one to grade five, The weighted fusion calculation result of grade five is greater than or equal to 0.95, the corresponding sphericity is greater than or equal to 0.93, the defect ratio is less than or equal to 0.1%, and the purity is greater than or equal to 99.5%; The weighted fusion calculation result of grade four is greater than or equal to 0.85, the corresponding sphericity is greater than or equal to 0.9, the defect ratio is less than or equal to 0.3%, and the purity is greater than or equal to 99%; The weighted fusion calculation result with a quality grade of level three is greater than or equal to 0.7, corresponding to a sphericity greater than or equal to 0.85, a defect ratio less than or equal to 0.5%, and a purity greater than or equal to 98%. The weighted fusion calculation result below level three of the quality grade is less than 0.7, which is unqualified spherical alumina.

[0033] Those skilled in the art can adjust the classification criteria of the quality grade according to the actual detection results.

[0034] In implementation, the method of the present invention constructs an intelligent detection system for multimodal fusion; through cross-modal data fusion of a high-resolution industrial camera, a laser scatterometer, and X-ray fluorescence spectroscopy, synchronous analysis of the morphological characteristics, particle size distribution, and chemical composition of spherical alumina is achieved; by establishing an association between the dynamic parameters of the flow velocity and the image acquisition frequency, modeling is realized. Due to the loss of key deformation frames caused by elastic deformation after particle collision, resulting in a decrease in the detection rate of surface defects, or due to the introduction of a large number of repeated images when the fluidity is stable, increasing the computational burden and introducing noise, and the displacement measurement window is too long, resulting in the problem of being unable to capture the transient switching of particle stagnation-flow, and the problem of decreased modeling accuracy caused by missed detection of abnormal data during the modeling stage. By adjusting the shooting frequency in real time according to the instantaneous velocity change amount before and after the collision, the effectiveness of image acquisition is improved. By adjusting the audit interval number of the data, the misjudgment of a single index is reduced, and the modeling accuracy is improved. By adjusting the comparison type of the image data, the comprehensiveness of data acquisition is improved.

[0035] Specifically, the process of adjusting the comparison type of the image data includes: Under the same conditions, the flow velocity of spherical alumina samples of the same batch on the inclined slide is collected several times; Calculate the velocity variance according to the several flow velocities; If the velocity variance is greater than or equal to a preset first variance, it is determined that the accuracy of model establishment does not meet the requirements; If the velocity variance is greater than or equal to a preset second variance, it is determined that the comprehensiveness of data acquisition does not meet the requirements, and the number of comparison types of the collected image data is increased; Wherein, the preset first variance is less than the preset second variance; the number of comparison types of the collected image data is positively correlated with the velocity variance.

[0036] Specifically, a number of industrial cameras are equidistantly arranged at the inclined edge of the inclined slide.

[0037] Specifically, the spherical alumina samples of the same batch are several equal-volume spherical alumina samples taken from the spherical alumina of a production batch.

[0038] Specifically, the velocity variance is the variance of the instantaneous velocities of several spherical alumina samples of equal volume flowing through the middle position of the inclined plane of the same inclined slide at one time. The calculation process of the variance is a well-known technical means to those skilled in the art and will not be elaborated herein.

[0039] Specifically, the instantaneous velocity is obtained by an industrial camera set at the middle position of the inclined plane of the inclined slide.

[0040] Specifically, the same conditions include that the volumes of the spherical alumina samples of the same batch for conducting the fluidity test are the same, the inclination angles of the inclined slides are the same, and the ambient temperatures of the fluidity test are the same.

[0041] Specifically, the same conditions also include that the vertical heights of the falls on the inclined slides are the same and the surface humidities on the inclined slides are the same.

[0042] Specifically, the comparison types of the image data include the minimum particle size value, the maximum particle size value, the median particle size, the particle size cumulative distribution, the specific surface area, the bulk density, and the particle shape.

[0043] Specifically, under the condition that the inclination angle of the inclined slide is 10°, the generally preset value range of the first variance is [0.4 m² / s², 0.6 m² / s²], and the generally preset value range of the second variance is [1 m² / s², 1.5 m² / s²].

[0044] Preferably, the preferred embodiment of the preset first variance is 0.5 m² / s², and the preferred embodiment of the preset second variance is 1.2 m² / s².

[0045] In implementation, when the difference between the velocity variance and the preset second variance is within 0.01 m² / s², the number of comparison types of the image data increases by 4. If the difference between the velocity variance and the preset second variance exceeds 0.01 m² / s², the increase in the number of comparison types and the difference between the velocity variance and the preset second variance satisfy the linear relationship N = 3×D(v), where N is the number of comparison types and D(v) is the ratio of the difference between the velocity variance and the preset second variance to 0.01 m² / s². For example, if the difference between the velocity variance and the preset second variance is 0.03 m² / s² and the current number of image comparison types is 20, the number of image comparison types increases to 20 + 3×0.03 m² / s² / 0.01 m² / s² = 29.

[0046] In implementation, by setting a first preset variance and a second preset variance in the method of the present invention, since only bright-field imaging is used, surface micro-cracks in the dark field and grain boundary anomalies in polarized light are prone to missed detection problems. For example, inconsistent interference of surrounding structures on image acquisition due to different acquisition angles, and the missing features caused by missed detection increase the misjudgment rate of the model, which in turn leads to the inability to match the morphology and composition data of the same particle, and the accuracy of multimodal fusion decreases. By increasing the comparison types, the accuracy and comprehensiveness of image acquisition are improved.

[0047] Specifically, the process of adjusting the audit interval number of image data includes: If the velocity variance is greater than or equal to the preset first variance and less than or equal to the preset second variance, it is preliminarily determined that the accuracy of the data participating in modeling does not meet the requirements; Obtain the output level of the model and the actual quality level of spherical alumina; According to the level difference between the output level of the model and the actual quality level of spherical alumina; If the level difference is greater than the preset level difference, it is secondarily determined that the accuracy of the data participating in modeling does not meet the requirements, and the audit interval number of the data is increased; Wherein, the audit interval number of the data is positively correlated with the level difference.

[0048] Specifically, auditing is to examine whether the image data is correct, for example, eliminating or marking redundant and incorrect image data through data cleaning and verification rules; the audit interval number is the number of images between two audits, for example, an audit is performed every time 100 new images are added.

[0049] Specifically, the level difference is the absolute value of the difference between the output level of the model and the actual quality level of spherical alumina.

[0050] Specifically, under the condition that the pixel size of the microscopic image of the spherical alumina sample collected by a 12-bit industrial camera is 3.45 μm, the preset level difference is 1 level.

[0051] In implementation, if the difference between the level difference and the preset level difference is greater than or equal to 1 level, the audit interval number of the data is increased according to the ratio result of the difference between the level difference and the preset level difference to 0.5. For example, if the difference between the level difference and the preset level difference is 2 levels and the current audit interval number of the data is 6, the audit interval number of the data is increased to 6 + 2 / 0.5 = 10.

[0052] In implementation, by setting a preset level difference amount, for the micro-displacement at the moment of collision, such as elastic deformation that cannot be accurately captured, and for the misdetection of components caused by XRF drift, by increasing the number of audit intervals, the problem of the accuracy decline of the model output result caused by the misjudgment of a single index is reduced, and the improvement of the model accuracy is achieved.

[0053] Specifically, the process of determining the shooting frequency for collecting the image data according to the change amount of the instantaneous velocity of the spherical alumina sample collision includes: Obtain the first displacement amount within the unit fluidity test duration before the collision and the second displacement amount within the unit fluidity test duration after the collision of the alumina that has a collision respectively; Calculate the change amount of the instantaneous velocity according to the first displacement amount and the second displacement amount; If the change amount of the instantaneous velocity is greater than the preset change amount, it is determined that the real-time performance of image acquisition does not meet the requirements, and the shooting frequency for collecting the image data is increased.

[0054] Specifically, the change amount of the instantaneous velocity is the difference between the instantaneous rate before the collision and the instantaneous rate after the collision.

[0055] Specifically, under the condition that the resolution of the industrial camera is 1μm / px, the general value range of the preset change amount is [0.1m / s, 0.5m / s], and the preferred embodiment of the preset change amount is 0.3m / s.

[0056] In implementation, when the difference between the change amount of the instantaneous velocity and the preset change amount is within 0.1m / s, the shooting frequency of the industrial camera is increased by 100fps. When the difference between the change amount of the instantaneous velocity and the preset change amount exceeds 0.1m / s, for every 0.1m / s exceeded, the shooting frequency of the industrial camera is increased by 120fps.

[0057] Specifically, the instantaneous rate before the collision is the ratio of the first displacement amount to the unit fluidity test duration; the instantaneous rate after the collision is the ratio of the second displacement amount to the unit fluidity test duration.

[0058] In implementation, by setting a preset change amount, due to the loss of key deformation frames caused by elastic deformation after particle collision, resulting in a decrease in the surface defect detection rate, or due to a large number of repeated images being introduced when the fluidity is stable, increasing the calculation burden and introducing noise, resulting in delays or errors in image acquisition, and further resulting in an iterative increase in the error amount of the participating data in the initial stage of modeling. By increasing the shooting frequency, an increase in modeling accuracy is achieved.

[0059] Specifically, the process of determining the unit fluidity test duration includes: Obtain the passing moments of the marked particle flows of spherical alumina samples at the ends of several inclined chutes; Calculate the standard deviation of the time intervals of several said passing moments; If the standard deviation of the time intervals is greater than or equal to a preset first standard deviation and less than a preset second standard deviation, then the unit fluidity test duration is set to a first-level duration; If the standard deviation of the time intervals is greater than or equal to the preset second standard deviation and less than a preset third standard deviation, then the unit fluidity test duration is set to a second-level duration; If the standard deviation of the time intervals is greater than or equal to the preset third standard deviation, then the unit fluidity test duration is set to a third-level duration.

[0060] Specifically, the marked particle flow of the spherical alumina sample is a spherical alumina particle group with a fluorescent dye or a magnetic coating, such as Fe3O4-coated, sprayed on the surface.

[0061] Specifically, the first-level duration is greater than the second-level duration, and the second-level duration is greater than the third-level duration.

[0062] Specifically, the time interval of the passing moment is the interval duration between the marked particle flows of spherical alumina samples passing through the ends of the inclined chutes in adjacent order.

[0063] Specifically, under the condition that the inclination angle of the inclined chute is 10°, the general value range of the preset first standard deviation is [0.08 s, 0.12 s], the general value range of the preset second standard deviation is [0.25 s, 0.4 s], and the general value range of the preset third standard deviation is [0.48 s, 0.55 s]; the preferred embodiment of the preset first standard deviation is 0.1 s, the preferred embodiment of the preset second standard deviation is 0.3 s, and the preferred embodiment of the preset third standard deviation is 0.5 s.

[0064] Specifically, under the condition that the inclination angle of the inclined chute is 10°, the general value range of the first-level duration is [100 ms, 120 ms], the general value range of the second-level duration is [40 ms, 60 ms], and the general value range of the third-level duration is [8 ms, 30 ms]; the preferred embodiment of the first-level duration is 100 ms, the preferred embodiment of the second-level duration is 50 ms, and the preferred embodiment of the third-level duration is 20 ms.

[0065] In implementation, the present invention determines the unit fluidity test duration, calculates the standard deviation of the particle flow time intervals, and establishes a three-level duration determination method. Through adaptive timing control, the best sampling rate can be maintained even when powder fluidity fluctuations are detected. By constructing a reverse duration adjustment mechanism, the sampling conditions are extended when the fluidity is poor to ensure data integrity and increase detection accuracy, while the duration is shortened when the fluidity is good to improve timeliness and reduce system power consumption.

[0066] So far, the technical solution of the present invention has been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, those skilled in the art can easily understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. An intelligent detection method for the quality of spherical alumina, characterized in that, Including: Collecting the microscopic images and chemical composition purity of spherical alumina samples; Segmenting and comparing the microscopic images to obtain the characteristic parameters of the spherical alumina samples so as to output the sphericity and defect ratio of the spherical alumina samples, wherein the characteristic parameters include the particle size and defect area of the spherical alumina samples; Inputting the sphericity, the defect ratio and the chemical composition purity of the spherical alumina samples into a multi-modal fusion model respectively, and outputting a quality grade evaluation result; Conducting a fluidity test on the spherical alumina samples, sequentially collecting, comparing and auditing the image data of the spherical alumina samples, and calculating the flow velocity of the spherical alumina samples; Determining the detection accuracy adjustment method according to the velocity variance of the flow velocities of a plurality of the spherical alumina samples, including adjusting the comparison type of the image data, or adjusting the audit interval number of the image data according to the grade difference between the output grade of the model and the actual quality grade of the spherical alumina; Wherein, the shooting frequency for collecting the image data is determined according to the instantaneous velocity change amount of the collision of the spherical alumina samples.

2. The intelligent detection method for the quality of spherical alumina according to claim 1, characterized in that, The process of adjusting the comparison type of the image data includes: Under the same conditions, collecting the flow velocities of the spherical alumina samples of the same batch on the inclined slideway for several times; Calculating the velocity variance according to the several flow velocities; If the velocity variance is greater than or equal to a preset first variance, it is determined that the accuracy of the model establishment does not meet the requirements; If the velocity variance is greater than or equal to a preset second variance, it is determined that the comprehensiveness of the data collection does not meet the requirements, and the number of comparison types of the collected image data is increased; Wherein, the preset first variance is less than the preset second variance; the number of comparison types of the collected image data is positively correlated with the velocity variance.

3. The intelligent detection method for the quality of spherical alumina according to claim 2, characterized in that, The same conditions include that the volumes of the spherical alumina samples of the same batch for conducting the fluidity test are the same, the inclination angle of the inclined slideway is the same, and the environmental temperature of the fluidity test is the same.

4. The intelligent detection method for the quality of spherical alumina according to claim 3, characterized in that, The process of adjusting the audit interval number of the image data includes: If the velocity variance is greater than or equal to the preset first variance and less than or equal to the preset second variance, it is preliminarily determined that the accuracy of the data participating in the modeling does not meet the requirements; Obtaining the output grade of the model and the actual quality grade of the spherical alumina; According to the grade difference between the output grade of the model and the actual quality grade of the spherical alumina; If the grade difference is greater than a preset grade difference, it is secondarily determined that the accuracy of the data participating in the modeling does not meet the requirements, and the audit interval number of the data is increased; Wherein, the audit interval number of the data is positively correlated with the grade difference.

5. The intelligent detection method for the quality of spherical alumina according to claim 4, characterized in that, The grade difference is the absolute value of the difference between the output grade of the model and the actual quality grade of the spherical alumina.

6. The intelligent detection method for the quality of spherical alumina according to claim 5, characterized in that, The process that the shooting frequency for collecting the image data is determined according to the instantaneous velocity change amount of the collision of the spherical alumina samples includes: Respectively obtaining a first displacement amount of the alumina that has collided within a unit fluidity test duration before the collision and a second displacement amount within a unit fluidity test duration after the collision; Calculate the change in the instantaneous velocity based on the first displacement and the second displacement; If the change in the instantaneous velocity is greater than a preset change, it is determined that the real-time performance of image acquisition does not meet the requirements, and the shooting frequency of acquiring the image data is increased.

7. The intelligent detection method for the quality of spherical alumina according to claim 6, characterized in that, The change in the instantaneous velocity is the difference between the instantaneous velocity rate before the collision and the instantaneous velocity rate after the collision.

8. The intelligent detection method for the quality of spherical alumina according to claim 7, characterized in that, The instantaneous velocity rate before the collision is the ratio of the first displacement to the unit fluidity test duration; the instantaneous velocity rate after the collision is the ratio of the second displacement to the unit fluidity test duration.

9. The intelligent detection method for the quality of spherical alumina according to claim 8, characterized in that, The determination process of the unit fluidity test duration includes: Obtain the passing moments of the marked particle flows of the spherical alumina samples at the ends of several inclined chutes; Calculate the standard deviation of the time intervals of several of the passing moments; If the standard deviation of the time intervals is greater than or equal to a preset first standard deviation and less than a preset second standard deviation, the unit fluidity test duration is set to a first-level duration; If the standard deviation of the time intervals is greater than or equal to the preset second standard deviation and less than a preset third standard deviation, the unit fluidity test duration is set to a second-level duration; If the standard deviation of the time intervals is greater than or equal to the preset third standard deviation, the unit fluidity test duration is set to a third-level duration.

10. The intelligent detection method for the quality of spherical alumina according to claim 9, characterized in that, The first-level duration is greater than the second-level duration, and the second-level duration is greater than the third-level duration.

Images