Material mixing uniformity evaluation method

Abstract

The invention provides a method for evaluating the uniformity of a mixed material. The method is used for solving the technical problem that an existing method for evaluating the uniformity of the mixed material cannot meet the requirement for evaluating the mixing uniformity of powder raw materials with different granularities and different densities. The method comprises the following steps: performing appropriate sampling at different positions of powder to be analyzed; photographing the surface of the sampled mixed material according to the granularity and specific characteristics of the raw material powder; selecting a proper subject matter in the photographed image, and sequentially carrying out binaryzation, subject matter marking, Delaunay triangulation, Voronoi image change and intercept point statistics on the photographed image; determining the mixing uniformity of a single sampling position according to an intercept point statistical result; counting the mixing uniformity of all sampling positions; and calculating the confidence degree of the mixing uniformity analysis result of all sampling positions by using distribution inspection, and determining the confidence degree of the judgment result. Compared with the prior art, the method has the advantages that the analysis speed is high, complicated chemical and physical pretreatment is not needed, real-time online monitoring can be realized, and the accuracy is relatively high.

Description

Technical Field

[0001] The invention belongs to the technical field of powder metallurgy and additive manufacturing, and in particular relates to a method for evaluating the uniformity of a mixture. Background Art

[0002] The mixing of powder materials refers to the process in which powder particles of different compositions and sizes change their movement speed or direction in the mixer, ultimately achieving uniform distribution of different particles in the mixer. The mixing of powder materials is widely used in the fields of pharmaceuticals, food, powder metallurgy, additive manufacturing, composite materials, abrasives and grinding tools. Due to the differences in particle size, density and surface properties of different raw material powders, there will be a certain degree of segregation of particles during the mixing process. The uniformity of the mixing directly determines the quality of the product, and its evaluation method is also crucial.

[0003] At present, there are many different methods for evaluating the uniformity of mixing. Different methods have their own advantages and limitations. High performance liquid chromatography (HPLC) has high sensitivity, but the cost of use is high; the conductivity method can be used to evaluate the mixing uniformity of conductive powders, but cannot be used for the analysis of insulating raw materials; thermal analysis (TGA / DSC) can provide information about the entire mixture, but insufficient offline sampling will introduce more errors; X-ray microchromatography can provide three-dimensional information of material powders, but cannot identify raw materials with similar structures and properties; electrical capacitance chromatography can be used in high temperature and high pressure environments, but is only suitable for powders with obvious changes in dielectric constants; acoustic emission spectroscopy can monitor the mixing effect online, but can only monitor the surface information of powders. Image analysis does not require complex pre-processing, has a fast analysis speed, and has the widest range of applications. At present, there are no reports on such applications.

[0004] The invention patent with application number 202211229029.8 relates to the technical field of high-speed mixers, and specifically discloses a method for evaluating the uniformity of a mixture based on a high-speed mixer. The method obtains a grayscale image of the surface image of the mixture; obtains the proportion of each raw material in the mixture according to the content of different raw materials to calculate the theoretical information entropy of the mixture, obtains the size of the tile window according to the content proportion, calculates the information entropy of the grayscale structure in each tile window, and obtains the grayscale texture uniformity of the grayscale image in combination with the theoretical information entropy; divides the grayscale image into multiple superpixels based on the grayscale value corresponding to each raw material, obtains the uniformity of the superpixel distribution of each raw material according to the Euclidean distance of the centroid between superpixels, and then obtains the uniformity of the comprehensive distribution of the raw materials; and takes the product of the uniformity of the grayscale texture and the uniformity of the comprehensive distribution of the raw materials as the final uniformity of the mixture. The above method improves the accuracy of the evaluation of the degree of mixture of the high-speed mixer and makes the evaluation result more credible. However, after the above patent grayscales the entire image, it focuses on the grayscale distribution of the entire image, which has a large amount of calculation and there is a certain interference in the grayscale of different substances. Summary of the invention

[0005] In view of the technical problem that the existing evaluation method of mixing uniformity cannot meet the evaluation requirements of mixing uniformity of powder raw materials with different particle sizes and densities, the present invention proposes a method for evaluating mixing uniformity. Based on the principle of image analysis, the Voronoi transformation and intercept statistical analysis method are used to quantitatively analyze the uniformity of the mixed powder. It is an online, non-destructive, efficient and accurate method for evaluating mixing uniformity. The characteristic components (subjects) in the mixture that the present invention focuses on are more focused, the overall calculation amount is small, and it is more suitable for the rapid calculation of large batches of images.

[0006] In order to achieve the above object, the technical solution of the present invention is implemented as follows: a method for evaluating the uniformity of a mixture, the steps of which are as follows:

[0007] Step 1: Sampling: Take appropriate amount of samples at different locations of the powder to be analyzed;

[0008] Step 2: Take photos: Take photos of the sampled mixed material surface according to the particle size and specific characteristics of the raw material powder;

[0009] Step 3: Image processing: Select appropriate objects in the photographed image, and perform binarization, object labeling, Delaunay triangulation, Voronoi diagram change, and interception statistics on the photographed image in sequence;

[0010] Step 4: Determination of the uniformity of the mixture at a single sampling location: Determine the uniformity of the mixture at a single sampling location based on the intercept point statistics;

[0011] Step 5: According to Step 2 to Step 4, the uniformity of the mixture at all sampling locations is counted;

[0012] Step 6: Result credibility assessment: Use distribution test to calculate the confidence of the mixture uniformity analysis results at all sampling locations to determine the credibility of the judgment results.

[0013] Preferably, the different positions in step 1 are distributed at different heights of the mixture, and a plurality of sampling positions are evenly distributed in the sampling surface corresponding to each position;

[0014] For large-particle powders with a particle size of 10 microns or more, an optical microscope is used to take pictures; for fine powders with a particle size of 10 microns or less, a scanning electron microscope is used to take pictures; for powders with fluorescent, ultraviolet, and infrared characteristics, corresponding cameras are used for photography; the magnification of the photography should be able to include as many targets as possible with clear resolution.

[0015] Preferably, the Delaunay triangulation is performed with the target object as a particle, so that the Delaunay triangle passes through all the target material points; the Delaunay triangulation method adopts an incremental construction method, a divide-and-conquer method or a convex hull method; the constructed Delaunay triangle satisfies: the circle formed by any two points does not contain other points; all generated triangles are "benign" and no angle is greater than or equal to 180 degrees.

[0016] Preferably, the Voronoi diagram is constructed using a Delaunay triangulation algorithm, and the implementation method of the Delaunay triangulation algorithm is as follows: selecting the outer centers of all Delaunay triangles and constructing a Voronoi diagram with the outer centers as intersection points; each region in the Voronoi diagram contains an object, and all points in the region are closer to the object than to other objects; points on the polygon boundary are at equal distances from points of the object that generated the polygon boundary, and the Voronoi polygon boundaries of adjacent figures take the original adjacent boundaries as subsets.

[0017] Preferably, the interception point statistics method is: using the interception line drawing method to divide a number of interceptions in the Voronoi diagram, the interceptions and the Voronoi diagram generate a series of interceptions, and counting the number of interceptions on each interception line;

[0018] The cutoffs are evenly distributed and cover most of the Voronoi diagram. The distribution of the cutoff points on each cutoff indicates the distribution of the target object. The uniformity of the cutoff point distribution can quantitatively analyze the uniformity of the mixed powder.

[0019] Preferably, the method for determining the mixing uniformity of a single sampling position in step 4 is: calculating the average number of intercept points on all intercepts on the Voronoi diagram, calculating the sample standard deviation of the number of intercept points on each intercept by the average number of intercept points, and dividing the sample standard deviation by the average number of intercept points to obtain the relative standard deviation; comparing the relative standard deviation with the standard threshold of the corresponding industry to determine whether the mixing process needs to be readjusted.

[0020] Preferably, the relative standard deviation The sample standard deviation of the number of intercept points on each intercept line The average intercept number In the formula, is the average number of intercept points in the Voronoi diagram; n is the number of intercepts; x i is the number of intercept points on each section line; s is the sample standard deviation of the number of intercept points on each section line; RSD is the relative standard deviation; the smaller the relative standard deviation RSD, the better the uniformity of the mixture.

[0021] In step 5, the relative standard deviation of each area is required to meet the requirements, indicating that the mixing uniformity of the entire system is good; otherwise, the mixing process, sampling and analysis need to be readjusted.

[0022] Preferably, the method for selecting the distribution test in step 6 is: when the sampling position m≤30, a t distribution test is used; when the sampling position m>30, a normal distribution is used.

[0023] Preferably, the method for implementing the confidence level of the t-distribution test analysis of the mixed powder uniformity analysis results at m sampling positions is: calculating the overall average intercept number

[0024] Calculate the overall mean cutoff number The average number of intercept points at each sampling position The difference

[0025] Calculate the overall standard deviation of a mixture system

[0026] Compute the 95% confidence interval for the t distribution:

[0027] Calculate the relative error of the statistics

[0028] Among them, n j is the number of intercepts in each sampling position; S j is the standard deviation of each sampling position j; t is the multiplication factor.

[0029] Preferably, the value of the multiplication factor t is obtained by looking up a t distribution table.

[0030] The beneficial effects of the present invention are as follows: Based on the principle of image analysis, the Voronoi transformation and intercept statistics are innovatively introduced to quantitatively analyze the uniformity of the mixed powder. Compared with the prior art, the present invention has a fast analysis speed, does not require complex chemical and physical pre-treatment, can be monitored online in real time, and has high accuracy. The present invention has the following advantages:

[0031] 1. Low cost: The image recognition method has a fast shooting speed, does not require complex chemical and physical pre-processing, and has lower costs;

[0032] 2. Good adaptability: The image magnification is adjustable and the light source is optional. It can evaluate the mixing uniformity of raw materials with different particle size ranges and different components, and has good adaptability;

[0033] 3. Accurate data: The combination of Voronoi transformation and intercept statistics quantifies the distribution of the standard particles in the image, and the data is accurate and reproducible. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0035] Figure 1 Flow chart of the mixing uniformity analysis process of the present invention.

[0036] Figure 2 Schematic diagram of a typical sampling position of the present invention, wherein (a) is a front view of a mixing barrel, (b) is a sampling surface 1 in (a), and (c) is a sampling surface 2 in (a).

[0037] Figure 3 This is an image of the mixing uniformity observed by the optical microscope method of the present invention.

[0038] Figure 4 This is an image of the mixing uniformity observed by scanning electron microscopy in the present invention.

[0039] Figure 5 The figure is a schematic diagram of the process of analyzing the uniformity of the mixture according to the present invention, wherein (a) is a binary image, (b) is a target object label, (c) is a Delaunay triangulation diagram, (d) is a schematic diagram of determining the circumcenter of the triangle, (e) is a Voronoi diagram change diagram, and (f) is a schematic diagram of intercept statistics.

[0040] Figure 6 This is a comparison chart of the ICP method used in the present invention to test the uniformity of the mixture. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] like Figure 1 As shown, a method for evaluating the uniformity of a mixture is shown, and the steps are as follows:

[0043] Step 1: Sampling. Take appropriate amounts of samples at different locations of the powder to be analyzed.

[0044] Specifically, the sampling positions and sampling amounts can refer to the Guidance for Industry: Powder Blends and Finished Dosage Units—Stratified In-Process Dosage Unit Sampling and Assessment issued by FDA in 2003 and the national standard GB / T 5918-2008 Determination of Mixing Uniformity of Feed Products. An appropriate number of sampling positions can be selected according to the amount of material. The sampling surfaces are generally distributed at different heights of the mixing barrel. The sampling positions of each sampling surface are evenly distributed in the sampling surface.

[0045] Step 2: Take a photo. Take a photo of the sampled mixed surface based on the particle size and specific characteristics of the raw material powder.

[0046] Specifically, the photography can be selected according to the particle size and specific characteristics of the powder to be analyzed. For example, for large-particle powder (particle size 10 microns and above), an optical microscope can be used for photography. For fine powder (particle size 10 microns and above), a scanning electron microscope can be used for photography. For powders with fluorescent, ultraviolet, infrared and other characteristics, corresponding cameras can be used for photography. The shooting magnification should be such that as many targets as possible can be included and the resolution should be clear.

[0047] Step 3: Image processing: Select appropriate objects in the photographed image, and perform binarization, object labeling, Delaunay triangulation, Voronoi transformation, and intercept statistics on the photographed image in sequence.

[0048] Specifically, the image processing can select the target object according to the specific characteristics of the mixed raw materials, such as the powder has typical morphology, typical color, typical fluorescence and other characteristics. Use software such as PhotoShop or corresponding algorithms (bimodal method, P parameter method, maximum inter-class variance method, maximum entropy threshold method, iterative method, etc.) to binarize the photographed image. The grayscale value of the pixel point on the binary image is set to 0 or 255, that is, the entire image presents an obvious visual effect of only black and white. The target object is the white point in the binary image. With the target object as the particle, Delaunay triangulation is performed (mathematical methods such as incremental construction method, divide and conquer method and convex hull method can be used) to make the Delaunay triangle pass through all the target material points. The constructed Delaunay triangle satisfies: the circle formed by any two points does not contain other points; all generated triangles are "benign", that is, no angle is greater than or equal to 180 degrees.

[0049] Select the outer center of all Delaunay triangles, which is the center of the circumscribed circle. Use the outer center as the intersection point to construct the Voronoi diagram using the Delaunay triangulation algorithm. Each region in the Voronoi diagram contains a target object, and all points in the region are closer to this target object than to other targets. The points on the polygon boundary are at the same distance from the target object points that generate this boundary. The Voronoi polygon boundary of the adjacent figure uses the original adjacent boundary as a subset. Divide several intercepts in the Voronoi diagram, and the intercepts and the Voronoi diagram generate a series of intercepts. The distribution of the intercepts should evenly cover most of the Voronoi image. You can refer to the American standard ASTM E112-2013 "Standard Test Methods for Determining Average Grain Size" for the intercept drawing method of grain size grading. The distribution of intercept points on each intercept line indicates the distribution of the target object. The uniformity of the intercept distribution can quantitatively analyze the uniformity of the mixed powder. If the number of intercept points on each intercept line is the same, it means that the material mixing is uniform. The position and number of intercept points can be obtained by cross product and collinearity detection method.

[0050] Step 4: Determination of mixing uniformity at a single sampling location: Determine the mixing uniformity of a single image (single sampling location) based on the intercept statistics.

[0051] Specifically, the intercept point statistics method can use formula 1-3 to calculate the mixing uniformity.

[0052]

[0053] In the formula, is the average number of intercept points in the Voronoi diagram; n is the number of intercepts; x i is the number of intercepts on each intercept line; s is the sample standard deviation of the number of intercepts on each intercept line; RSD is the relative standard deviation. The smaller the RSD, the better the uniformity of the mixing. Different RSD standards can be set for different industry needs. For example, the pharmaceutical and chemical industry can use RSD≤5% as the standard, and the powder metallurgy industry can use RSD≤10% as the standard. If the RSD exceeds the standard, the mixing process needs to be readjusted.

[0054] Step 5: Determine the uniformity of the overall mixture. According to steps 2 to 4, the uniformity of the mixture at all sampling locations is counted and macroscopically quantified.

[0055] Specifically, all sampling positions in step 1 are analyzed in steps 2 to 4 to obtain the relative standard deviation RSD of each sampling position. Generally, the RSD of each area is required to meet the requirements, which indicates that the mixing uniformity of the entire system is good. Otherwise, the mixing process, sampling, and analysis need to be readjusted.

[0056] Step 6: Result credibility assessment: Using the distribution test, calculate the confidence of the mixture uniformity analysis results at all sampling locations and determine the credibility of the judgment results.

[0057] Specifically, the confidence of the mixed powder uniformity analysis results at m sampling positions is analyzed, and the calculation formula is as follows:

[0058]

[0059] In the formula, is the overall average cutoff number; is the average number of intercept points at each sampling position; n j is the number of intercepts in each sampling position; m is the number of sampling positions; d j is the overall average cutoff number The average number of intercept points at each sampling position The difference of S T is the total standard deviation of the mixing system; S j is the standard deviation of each sampling position j; 95%CI is the 95% confidence interval of t distribution (when sampling position m≤30, t distribution test can be used; when m>30, normal distribution can be used); the value of multiplication factor t can be found in Table 1 from the t distribution table; %RA is the relative error of statistics. For most applications, the relative error %RA≤10% is considered valid. If the relative error %RA is too large, it is necessary to confirm whether the sampling position and sampling method are reasonable, and re-sample and analyze after adjustment.

[0060] Table 1 95% confidence internal multiplication factor of t distribution (partial)

[0061]

[0062] Example 1

[0063] A mixing uniformity evaluation method of the present invention is used to evaluate the mixing uniformity of a batch of main raw material powders with a particle size of 10-20 microns. The main process is shown in FIG. Figure 1 The specific evaluation plan is:

[0064] (1) Sampling. Samples were taken at 8 different locations of the powder to be analyzed. The sampling locations are shown in Figure 2 This method can obtain multiple samples at different planes and coordinates in the mixed system, and is highly representative.

[0065] (2) Take photos. A certain raw material powder A in the powder to be analyzed appears bright white under the backscattered electrons of the scanning electron microscope, while other raw material powders are mainly dark gray, so the backscattering mode is used for photography. The main particle size of powder A is 10-20 microns, so the magnification is 1600 times, see Figure 4 .

[0066] (3) Image processing. Powder A was selected as the target object; Photoshop software was used to binarize the image to increase the contrast between powder A and the surrounding background; the object markers in the image were manually selected; the incremental construction method was used to perform Delaunay triangulation so that the Delaunay triangle passed through all the target material points; the outer circle center of all Delaunay triangles was selected; a Voronoi diagram was constructed with the outer circle center as the intersection point so that each area in the Voronoi diagram contained a target object, and all points in the area were closer to this target object than to other targets; 8 line segments with 4 horizontal and 4 vertical lines were used as transversals to fill the Voronoi diagram, and then the number of transversals was counted to obtain Figure 5 .

[0067] (4) Determine the mixing uniformity of the sampling position. According to the intercept statistical results (Table 2) and formulas 1-3, determine the mixing uniformity of the sampling position. From the results, it can be seen that the relative standard deviation RSD of the distribution of powder A is 17.5%. If the relative standard deviation RSD ≤ 10% is taken as the qualified standard, the mixing uniformity of this batch of powder is insufficient. If the relative standard deviation RSD ≤ 20% is taken as the qualified standard, the mixing uniformity of this batch of powder is qualified.

[0068] Table 2 Cutoff point statistics

[0069]

[0070] (5) Determination of overall mixing uniformity. According to steps 2 to 4, the mixing uniformity of all 8 sampling positions is statistically analyzed, and the results are shown in Table 3. According to formulas 1-3, the overall average intercept number of the 8 sampling positions is 4.62, the total standard deviation is 0.53, and the total relative standard deviation RSD is 11.5%. This shows that under the current mixing process, the distribution of powder A at the microscopic level has certain fluctuations.

[0071] Table 3 Image analysis results of 8 sampling locations

[0072] Sampling location Position 1 Position 2 Location 3 Position 4 Location 5 Position 6 Position 7 Position 8 Average number of intercept points 4.75 4.61 4.66 4.57 4.58 4.62 4.60 4.59 Number of intercepts 8 8 8 8 8 8 8 8 Standard Deviation 0.83 0.55 0.46 0.46 0.49 0.39 0.38 0.53

[0073] (6) Result credibility assessment. Using the t-distribution test, calculate the confidence of the mixture uniformity analysis results at all sampling locations to determine the credibility of the judgment results. As shown in Table 1, the sampling location is 8 and the t value is 2.365. According to formula 4-8, the corresponding relative standard deviation 95% CI is 0.44 and the relative error %RA is 9.6%. The relative error %RA≤10%, indicating that the sampling method is effective and the uniformity analysis results are valid. If the relative error %RA is too large, it is necessary to confirm whether the sampling method is reasonable, and re-sample and analyze after adjustment.

[0074] Comparative Example 1

[0075] Ordinary optical microscopy was used to determine the uniformity of the raw material mixing in the examples. The raw material was magnified 50 times, and the optical photograph obtained was shown in Figure 3 . Figure 3 and Figure 4 The sampling position is the same, but the magnification is different, so it is not the same micro-area. Figure 3 It can be seen that under low magnification conditions, the overall color of the powder to be analyzed is uniform and the mixing effect is good. The particle size of the target (powder A) is 10-20 microns and is rose red. Under a 50x optical microscope, the target is relatively blurred and it is impossible to confirm whether it is agglomerated or not, and it is impossible to make a quantitative description.

[0076] Comparative Example 2

[0077] The ICP method (inductively coupled plasma spectroscopy) was used to test the mixing uniformity of the raw materials in the embodiment, and the results are shown in Figure 6 . From the results, it can be seen that the relative standard deviation RSD of the target (powder A) at the 10 sampling positions is 2.77%, which is relatively uniform macroscopically. However, the ICP method has many limitations. First, the ICP method requires the dissolution of the entire mixed powder. If part of the raw materials cannot be dissolved, filtration, digestion and other treatments are required, and the pre-treatment is complicated and time-consuming. Secondly, since ICP is mainly used for the analysis of medium and low concentrations (especially trace amounts) of substances, the measurement of high-concentration components requires a large proportion of dissolution and dilution, which will introduce large errors and is not conducive to powder uniformity testing. Thirdly, a large amount of raw material powder is required for ICP dissolution, and uniformity analysis at the microscopic level cannot be performed. Finally, due to the principle limitations of spectroscopy, there is mutual interference between the spectra of many elements, so special attention should be paid when testing such elements.

[0078] Example 1, Comparative Example 1 and Comparative Example 2 are analyses of the same batch of raw materials. By comparison, it can be seen that the scheme of the present invention has good adaptability and the analysis results can be quantified. Comparative Example 1 cannot meet the uniformity determination of fine powders and cannot be quantified. Although Comparative Example 2 can be quantified, the test is cumbersome and cannot be applied to the uniformity determination of raw materials at the microscopic level.

[0079] The present invention is based on the principle of image analysis, and innovatively introduces Voronoi transformation and intercept statistics to quantitatively analyze the uniformity of the mixed powder. The present invention has a fast analysis speed, does not require complex chemical and physical pre-treatment, can be monitored online in real time, and has high accuracy. The present invention ensures the uniformity of mixed powders through the entire process of scientific sampling, image conversion, single sampling point analysis, and macroscopic analysis; and determines the confidence of the overall sampling system and test results through distribution inspection. The full set of evaluation systems is perfect and has a wide range of applications. The present invention innovatively combines Voronoi transformation and intercept statistics, which can be used to zonal and quantitatively judge the mixing uniformity effects of different types of powders, with low analysis cost and fast speed.

[0080] The present invention protects a method for evaluating the uniformity of a mixture, including but not limited to the above embodiments.

[0081] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for evaluating mixing uniformity, characterized in that: The steps are as follows: Step 1: Sampling: Take appropriate amount of samples at different locations of the powder to be analyzed; Step 2: Take photos: Take photos of the sampled mixed material surface according to the particle size and specific characteristics of the raw material powder; Step 3: Image processing: Select appropriate objects in the photographed image, and perform binarization, object labeling, Delaunay triangulation, Voronoi diagram change, and interception statistics on the photographed image in sequence; Step 4: Determination of the uniformity of the mixture at a single sampling location: Determine the uniformity of the mixture at a single sampling location based on the intercept point statistics; Step 5: According to Step 2 to Step 4, the uniformity of the mixture at all sampling locations is counted; Step 6: Result credibility assessment: Use distribution test to calculate the confidence of the mixture uniformity analysis results at all sampling locations to determine the credibility of the judgment results.

2. The method for evaluating mixing uniformity according to claim 1, characterized in that: The different positions in step 1 are distributed at different heights of the mixture, and a number of sampling positions are evenly distributed in the sampling surface corresponding to each position; For large-particle powders with a particle size of 10 microns or more, an optical microscope is used to take pictures; for fine powders with a particle size of 10 microns or less, a scanning electron microscope is used to take pictures; for powders with fluorescent, ultraviolet, and infrared characteristics, corresponding cameras are used for photography; the magnification of the photography should be able to include as many targets as possible with clear resolution.

3. The method for evaluating mixing uniformity according to claim 1 or 2, characterized in that: The Delaunay triangulation is to perform Delaunay triangulation with the target object as a particle, so that the Delaunay triangle passes through all the target material points; the Delaunay triangulation method adopts the incremental construction method, the divide-and-conquer method or the convex hull method; the constructed Delaunay triangle satisfies: the circle formed by any two points does not contain other points; all generated triangles are "benign", and no angle is greater than or equal to 180 degrees.

4. The method for evaluating mixing uniformity according to claim 1 or 2, characterized in that: The construction method of the Voronoi diagram is a Delaunay triangulation algorithm, and the implementation method of the Delaunay triangulation algorithm is: select the outer circle center of all Delaunay triangles, and construct a Voronoi diagram with the outer circle center as the intersection point; each area in the Voronoi diagram contains a target object, and the distance between all points in the area and this target object is closer than the distance between all points in the area and other targets; the distance between the points on the polygon boundary and the points of the target object that generates this polygon boundary is equal, and the Voronoi polygon boundary of the adjacent figure takes the original adjacent boundary as a subset.

5. The method for evaluating mixing uniformity according to claim 4, characterized in that: The method for interception point statistics is: using the interception line drawing method to divide several interceptions in the Voronoi diagram, the interceptions and the Voronoi diagram generate a series of interceptions, and counting the number of interceptions on each interception line; The cutoffs are evenly distributed and cover most of the Voronoi diagram. The distribution of the cutoff points on each cutoff indicates the distribution of the target object. The uniformity of the cutoff point distribution can quantitatively analyze the uniformity of the mixed powder.

6. The method for evaluating mixing uniformity according to claim 5, characterized in that: The method for determining the mixing uniformity of a single sampling position in step 4 is: calculating the average number of intercept points on all intercepts on the Voronoi diagram, calculating the sample standard deviation of the number of intercept points on each intercept by the average number of intercept points, and dividing the sample standard deviation by the average number of intercept points to obtain the relative standard deviation; using the relative standard deviation to compare with the standard threshold of the corresponding industry to determine whether the mixing process needs to be readjusted.

7. The method for evaluating mixing uniformity according to claim 6, characterized in that: The relative standard deviation The sample standard deviation of the number of intercept points on each intercept line The average intercept number In the formula, is the average number of intercept points in the Voronoi diagram; n is the number of intercepts; x i is the number of intercept points on each intercept line; s is the sample standard deviation of the number of intercept points on each intercept line; RSD is the relative standard deviation; the smaller the relative standard deviation RSD, the better the uniformity of the mixture; In step 5, the relative standard deviation of each area meets the requirements, indicating that the mixing uniformity of the entire system is good; Otherwise, the mixing process, sampling and analysis need to be readjusted.

8. The method for evaluating mixing uniformity according to any one of claims 4 to 7, characterized in that: The selection method of the distribution test in step 6 is: when the sampling position m≤30, the t distribution test is adopted; when the sampling position m>30, the normal distribution is adopted.

9. The method for evaluating mixing uniformity according to claim 8, characterized in that: The implementation method of the t distribution test to analyze the confidence of the mixed powder uniformity analysis results of m sampling positions is: calculate the overall average intercept number Calculate the overall mean cutoff number The average number of intercept points at each sampling position The difference Calculate the overall standard deviation of a mixture system Compute the 95% confidence interval for the t distribution: Calculate the relative error of the statistics Among them, n j is the number of intercepts in each sampling position; S j is the standard deviation of each sampling position j; t is the multiplication factor.

10. The method for evaluating mixing uniformity according to claim 9, characterized in that: The value of the multiplication factor t is obtained by looking up a t distribution table.

Images