Multi-thread processing method for intelligently optimizing working parameters of blender mixer

Abstract

The invention relates to the technical field of mixing processing, and discloses a multi-thread processing method for intelligently optimizing working parameters of a mixer. Comprising the following steps: obtaining initial stirring time according to a proportional relation, mixing to-be-mixed materials according to the initial stirring time to obtain an initial mixture, obtaining particle characteristics according to a real-time surface image, and calculating a mixing influence degree value based on real-time detection data, inputting the proportional relation, the particle characteristics and the mixing influence degree value into a pre-constructed dose generation model to obtain an antistatic dose, and determining a mixing time interval and a mixing rotating speed interval according to the proportional relation and the antistatic dose, determining an optimal mixing time and an optimal mixing rotating speed based on the mixing time interval, the mixing rotating speed interval and an optimization algorithm; the combination of the mixing time and the mixing rotating speed after optimization can reduce the generation of static electricity to the greatest extent, thereby avoiding the agglomeration or non-uniformity of the mixture, and further improving the overall quality and consistency of the product.

Description

Technical Field

[0001] The present invention relates to the technical field of material mixing processing, and more specifically, to a multi-thread processing method for intelligently optimizing working parameters of a material mixer. Background Art

[0002] A mixer is a mechanical device used to mix different materials evenly. It effectively mixes multiple materials together through mechanical force to ensure that the final product meets the requirements of uniformity and consistency. The equipment is widely used in chemical, pharmaceutical, food, building materials, metallurgy and other industries. It is mainly used for mixing materials such as powders and particles. Among these applications, the production of glass powder is particularly typical. Glass powder is an amorphous hard particle powder prepared by mixing oxides such as silicon dioxide, aluminum oxide, and magnesium oxide. It has excellent wear resistance, corrosion resistance and mechanical strength, and is low in cost. It is very suitable for large-scale industrial production. Therefore, the mixer plays a key role in the preparation process of glass powder, ensuring its production efficiency and product quality.

[0003] In the process of preparing glass powder by a mixer, some problems are usually encountered and corresponding optimization is required. Most of the existing technologies focus on optimizing the proportion of each component in the process of preparing glass powder, or improving the preparation process itself, but rarely involve optimizing the working parameters in the preparation process. For example, the Chinese patent with publication number CN105198211A provides a low-melting-point glass powder and a preparation method thereof. The patent sets the proportion of each component in the glass powder to generate glass powder with the advantages of low glass transition temperature and adjustable thermal expansion coefficient. The inventors found through in-depth research and practical application of the existing technology that there are at least the following defects in the optimization process of glass powder preparation:

[0004] Since glass powder is prepared by mixing oxide particle powders such as silicon dioxide, aluminum oxide, and magnesium oxide, these oxide particles are prone to generate static electricity due to friction during the mixing process. The generation of static electricity will cause the particle powders to attract each other, forming agglomerations or lumps, thereby affecting the uniformity of the mixing; in addition, the accumulation of static electricity is closely related to the working parameters in the preparation process, and the existing technology has done little to optimize these parameters to reduce the generation of static electricity. Therefore, it is necessary to further optimize the working parameters to effectively reduce the static electricity problem in the mixing process.

[0005] In view of this, the present invention proposes a multi-thread processing method for intelligent optimization of mixer working parameters to solve the above problems. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, the present invention provides a multi-thread processing method for intelligently optimizing the working parameters of a mixer.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] In the first aspect, a multi-thread processing method for intelligently optimizing the working parameters of a mixer includes:

[0009] Obtaining the ratio relationship between the sub-powder types in the material to be mixed, inputting the ratio relationship into a pre-built time determination model to obtain an initial stirring time, and mixing the material to be mixed according to the initial stirring time to obtain an initial mixture;

[0010] Acquire a real-time surface image and real-time detection data of the initial mixture, acquire particle characteristics according to the real-time surface image, the particle characteristics at least include particle size characteristics and shape characteristics, and calculate the mixing influence value based on the real-time detection data;

[0011] Input the proportional relationship, particle characteristics and mixing influence value into the pre-built dose generation model to obtain the antistatic dose, and determine the mixing time interval and mixing speed interval according to the proportional relationship and the antistatic dose;

[0012] The optimal mixing time and the optimal mixing speed are determined based on the mixing time interval, the mixing speed interval and the optimization algorithm.

[0013] Furthermore, the particle size feature is the average value of the particle size of the powder particles, and the shape feature is the average roundness of the particles. The method for obtaining the particle size feature according to the real-time surface image includes:

[0014] Any corner point of the real-time surface image is taken as the origin of the detection coordinate system, and the two edges connected by the corner point are taken as the two axes of the detection coordinate system. In the detection coordinate system, R groups of detection points are generated by the Markov random walk method. The equivalent diameters of the particles corresponding to the R group of detection points are extracted according to the edge detection algorithm, and the average particle size of the equivalent diameters of the R group of particles is calculated, and the average particle size is used as the particle size feature.

[0015] Furthermore, the method for extracting the equivalent diameter of particles according to the edge detection algorithm includes:

[0016] The effective area of ​​the particles corresponding to the detection points in group R is extracted according to the edge detection algorithm, and the equivalent diameter of the particles is calculated according to the effective area of ​​the particles;

[0017] Methods for calculating the equivalent diameter of particles based on the effective area of ​​the particles include:

[0018]

[0019] Where EDP is the equivalent diameter of the particle, EAP is the effective area of ​​the particle, and π is a constant.

[0020] Furthermore, the method for acquiring shape features according to the real-time surface image includes:

[0021] The effective perimeter of the particles corresponding to the detection points of group R is extracted according to the edge detection algorithm, the roundness of the particles is obtained according to the effective area and the effective perimeter of the particles, and the average roundness of the particles of group R is taken as the shape feature;

[0022] Methods for obtaining particle roundness based on particle effective area and particle effective circumference include:

[0023]

[0024] Where ROP is the roundness of the particle and EFP is the effective circumference of the particle.

[0025] Furthermore, the method for constructing the time determination model includes:

[0026] A first sample data set is obtained, wherein the first sample data set includes a historical proportion relationship and a historical initial stirring time. The first sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical proportion relationship in the sample training set is used as input data of the regression network, and the historical initial stirring time in the sample training set is used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting the real-time initial stirring time. The initial regression network is tested using the sample test set, and the initial regression network that satisfies a preset error value is output as a time determination model.

[0027] Furthermore, the real-time detection data includes contact potential difference value, leakage current value and electric field strength value, and the mixed influence value is calculated according to the contact potential difference value, leakage current value and electric field strength value.

[0028] Further, the method for determining the mixing time interval according to the proportional relationship and the antistatic dosage includes:

[0029] Based on the proportional relationship, multiply the corresponding preset single-type mixing time, and then sum them up to obtain the first mixing time. Calculate the dose ratio based on the antistatic dose and the preset standard dose, multiply the first mixing time by the dose ratio to obtain the second mixing time, and use the first mixing time and the second mixing time as the endpoint values ​​of the mixing time interval respectively.

[0030] Furthermore, the optimization algorithm is an ant colony algorithm, and the method for determining the optimal mixing time and the optimal mixing speed according to the ant colony algorithm includes:

[0031] S401: Preset the number of ants to be M, the pheromone intensity to be τ0, the volatility coefficient to be ρ, the importance coefficient of the pheromone to be α, and the importance coefficient of the heuristic information to be β;

[0032] S402: discretizing the mixed time interval and the mixed speed interval into N nodes, and the ant selects a path between the N nodes;

[0033] S403: Taking the minimization of the mixing influence value as the optimization goal, heuristic information is calculated for each node, and the heuristic information is the inverse value of the mixing time and the mixing speed, that is:

[0034]

[0035] Where η ij is the heuristic information of the ant from node i to node j, MIV ij The mixed influence value of the combination of mixing time and mixing speed selected by the ant from node i to node j;

[0036] S404: The ants select a path based on the pheromone concentration and heuristic information. The probability of going from node i to node j is:

[0037]

[0038] In the formula, p ij is the probability of an ant moving from node i to node j, τ ij is the pheromone strength from node i to node j, τ ik is the pheromone strength from node i to node k, η ik is the heuristic information from node i to node k, and allowed represents the set of paths that the ant can choose;

[0039] S405: Each ant selects a path and determines the mixed time and mixed speed in turn until the ant walks through all nodes and constructs a complete combined solution of the mixed time and mixed speed. The mixed influence value MIV is calculated for the combined solution constructed by each ant, and the combined solution with the smallest mixed influence value in the current iteration is recorded.

[0040] S406: updating the pheromones on the path;

[0041] S407: Repeat S403-S406 until the maximum number of iterations is reached, then stop the iterations and output the mixing time and mixing speed corresponding to the path with the highest pheromone concentration as the optimal mixing time and the optimal mixing speed.

[0042] Furthermore, the method for updating the pheromone on the path includes:

[0043]

[0044] In the formula, τ ij (t+1) is the pheromone strength from node i to node j in the t+1th iteration, τ ij (t) is the pheromone strength from node i to node j in the tth iteration, Δτ ijis the total pheromone increment left by all ants on the path from node i to node j in the tth iteration;

[0045] Δτ ij The definitions include:

[0046]

[0047] Where M is the total number of ants, is the incremental pheromone of the f-th ant from node i to node j;

[0048] The definitions include:

[0049]

[0050] Where MIV f is the mixed influence value corresponding to the f-th ant passing through node i to node j, and Q is a constant greater than 0.

[0051] In a second aspect, a computer-readable storage medium stores a computer program, and when the computer program is executed, the multi-threaded processing method for intelligently optimizing the working parameters of the mixer is implemented.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] The present invention first obtains the proportional relationship between each sub-powder type in a to-be-mixed material, obtains an initial stirring time according to the proportional relationship, mixes the to-be-mixed material according to the initial stirring time to obtain an initial mixture, then obtains a real-time surface image and real-time detection data of the initial mixture, obtains particle characteristics according to the real-time surface image, calculates a mixing influence value based on the real-time detection data, inputs the proportional relationship, particle characteristics and mixing influence value into a pre-constructed dose generation model, obtains an antistatic dose, determines a mixing time interval and a mixing speed interval according to the proportional relationship and the antistatic dose, determines an optimal mixing time and an optimal mixing speed based on the mixing time interval, the mixing speed interval and an optimization algorithm, and the optimized mixing time and mixing speed combination can minimize the generation of static electricity, avoid agglomeration or unevenness of the mixture, and thereby improve the overall quality and consistency of the product. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A flow chart of a multi-thread processing method for intelligent optimization of working parameters of a mixer in the present invention;

[0055] Figure 2 It is a structural schematic diagram of a multi-thread processing system for intelligent optimization of working parameters of a mixer in the present invention;

[0056] Figure 3 is a schematic diagram of the detection coordinate system in the present invention;

[0057] Figure 4 It is an enlarged schematic diagram of the detection point in the present invention;

[0058] Figure 5 A schematic diagram of a computer-readable storage medium in the present invention.

[0059] Reference numerals:

[0060] 10. Real-time surface image; 20. Detection coordinate system; 30. Detection point; 40. Complete powder particles; 50. Incomplete powder particles. DETAILED DESCRIPTION

[0061] 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.

[0062] Example 1

[0063] See also Figure 1 As shown, this embodiment discloses a multi-threaded processing method for intelligently optimizing the working parameters of a mixer, including:

[0064] S10: obtaining a ratio relationship between various sub-powder types in the material to be mixed, inputting the ratio relationship into a pre-built time determination model, obtaining an initial stirring time, and mixing the material to be mixed according to the initial stirring time to obtain an initial mixed material;

[0065] In this embodiment, the material to be mixed can be glass powder, and the sub-powder types include at least silicon dioxide, aluminum oxide and magnesium oxide. The above-mentioned proportional relationship refers to the component proportional relationship between each sub-powder type. It can be understood that the proportional relationship between each sub-powder type is predetermined, and each sub-powder is passed into the mixer according to the proportional relationship to form the material to be mixed.

[0066] The construction method of the time determination model includes:

[0067] A first sample data set is obtained, wherein the first sample data set includes a historical proportion relationship and a historical initial stirring time. The first sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical proportion relationship in the sample training set is used as input data of the regression network, and the historical initial stirring time in the sample training set is used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting the real-time initial stirring time. The initial regression network is tested using the sample test set, and the initial regression network that satisfies a preset error value is output as a time determination model. The initial regression network is preferably a deep neural network model.

[0068] It should be added that due to the different types of sub-powders, the density and friction coefficient of the granular powders are also different, and these differences will directly affect the setting of the initial stirring time, and then different proportions will affect the length of the stirring time. For example, when the sub-powder types include silica, alumina and magnesium oxide, the density of silica granular powder is lower, but the friction coefficient is higher; while the density of magnesium oxide granular powder is higher and the friction coefficient is lower. Powders with higher density are generally more likely to settle, so longer stirring times are required to ensure their uniform distribution. On the contrary, powders with higher friction coefficients are more likely to generate static electricity during the mixing process, so they should not be stirred for a long time to avoid agglomeration or agglomeration. For example, when the ratio of silica, alumina and magnesium oxide is 5:2:3, the initial stirring time is relatively short due to the higher proportion of silica and the larger friction coefficient, and when the ratio is 3:2:5, the initial stirring time is correspondingly longer due to the increased proportion of magnesium oxide and its high density characteristics requiring a longer time to ensure uniform distribution.

[0069] The method of mixing the materials to be mixed according to the initial stirring time to obtain the initial mixed material includes:

[0070] The initial stirring time and the preset initial stirring speed are input into the mixer, and the mixer mixes the materials to be mixed to obtain the initial mixture.

[0071] The mixer mentioned above can be a V-type mixer or a ribbon mixer, etc., and the initial stirring speed is pre-set by those skilled in the art, and there are also corresponding parameters that are pre-set, such as the air humidity inside the mixer. The air humidity can be pre-set at 50%. The air humidity in this embodiment usually refers to relative humidity, that is, the ratio of the actual amount of water vapor in the air to the maximum amount of water vapor that the air can accommodate at that temperature, expressed as a percentage.

[0072] In this embodiment, the initial stirring stage helps to initially evenly disperse different types of powders and reduce the problem of uneven mixing that may occur in the subsequent formal stirring process. In particular, when there are large differences in particle density and friction coefficient, the initial stirring can effectively avoid local uneven mixing caused by these differences. At the same time, powders with a higher friction coefficient (such as silica) are prone to generate static electricity during the stirring process, causing the particles to attract each other, agglomerate or agglomerate. By reasonably setting the initial stirring time, the friction time can be effectively controlled, static electricity accumulation can be reduced, and large-scale agglomeration problems can be avoided in formal stirring. In addition, setting the initial stirring time according to the proportional relationship of the sub-powders can avoid the stirring time being too long or too short, thereby reducing unnecessary energy consumption and improving stirring efficiency.

[0073] S20: obtaining a real-time surface image and real-time detection data of the initial mixture, obtaining particle characteristics according to the real-time surface image, the particle characteristics at least including particle size characteristics and shape characteristics, and calculating a mixing influence value based on the real-time detection data;

[0074] Among them, the real-time surface image can be obtained through the camera unit in the mixer, and the camera unit is a camera installed in the mixer. The particle size feature mentioned above refers to the average particle size of the powder particles, and the shape feature refers to the average roundness of the particles. The particle roundness is usually defined as the square of the circumference of the particle divided by its area multiplied by a constant, and the constant can be 4π.

[0075] Methods for obtaining particle size characteristics based on real-time surface images include:

[0076] Any corner point of the real-time surface image is taken as the origin of the detection coordinate system, and the two edges connected by the corner point are taken as the two axes of the detection coordinate system. In the detection coordinate system, R groups of detection points are generated by the Markov random walk method. The equivalent diameters of the particles corresponding to the R group of detection points are extracted according to the edge detection algorithm, and the average particle size of the equivalent diameters of the R group of particles is calculated, and the average particle size is used as the particle size feature.

[0077] In this embodiment, the corner point refers to a vertex of the real-time surface image, and the two sides connected by the corner point refer to the image boundaries extending from the corner point. Figure 3 As shown, Figure 3 A real-time surface image 10 , a detection coordinate system 20 and detection points 30 are shown.

[0078] Methods for extracting particle equivalent diameters based on edge detection algorithms include:

[0079] The effective area of ​​the particles corresponding to the R group of detection points is extracted according to the edge detection algorithm, and the equivalent diameter of the particles is calculated based on the effective area of ​​the particles.

[0080] Methods for calculating the equivalent diameter of particles based on the effective area of ​​the particles include:

[0081]

[0082] Where EDP is the equivalent diameter of the particle, EAP is the effective area of ​​the particle, and π is a constant.

[0083] In the above, the effective area of ​​the particle refers to the complete particle area within the detection point, such as Figure 4 As shown, Figure 4 The detection point 30, the complete powder particle 40 and the incomplete powder particle 50 are shown, then the area of ​​the complete powder particle 40 can be used as the effective area of ​​the particle, the above-mentioned edge detection algorithm can be Canny, Sobel, etc., and the above-mentioned Markov random walk method refers to a random process model based on Markov chain, which is used to simulate randomly changing systems or processes. This method is often used in calculations to simulate random paths, predictions and other uncertain system analyses. The basic idea of ​​the Markov random walk method is that the next state of the system depends only on the current state and is independent of the previous state (that is, it has "memorylessness").

[0084] It can be understood that in this embodiment, a real-time surface image of the initial mixture is obtained. The purpose of doing so is that the initial mixture has been preliminarily stirred and the distribution of particles in space is relatively uniform. At this time, obtaining a surface image can more accurately reflect the actual mixing state of each component, thereby helping to further adjust the mixing parameters and ensure the final mixing uniformity. Although the edge detection algorithm can effectively identify the edges of particles, directly using edge detection to generate particle size features may be affected by factors such as image noise and blurred particle boundaries. By generating detection points and calculating the effective area of ​​the particles, and then calculating the equivalent diameter of the particles based on the effective area, the actual size of the particles can be more accurately reflected. The combined use of Markov random walk method, edge detection algorithm and equivalent diameter calculation method can make the system more robust, and can stably generate reliable particle size features under various environments and conditions, thereby improving the reliability of mixing process monitoring.

[0085] Similarly, methods for obtaining shape features based on real-time surface images include:

[0086] The effective circumference of the particles corresponding to the R group detection points is extracted according to the edge detection algorithm, and the particle roundness is obtained according to the effective area and the effective circumference of the particles. The average value of the roundness of the R group particles is used as the shape feature.

[0087] Methods for obtaining particle roundness based on particle effective area and particle effective circumference include:

[0088]

[0089] Where ROP is the roundness of the particle and EFP is the effective circumference of the particle.

[0090] In the above, the effective circumference of the particle refers to the circumference of the complete particle within the detection point. It can be understood that whether it is the above-mentioned particle equivalent diameter or particle roundness, it refers to the average value of the diameters of several particle powders within a detection point or the average value of the roundness of the particle powder. The average value of one detection point is calculated first, and then the average value within R detection points is calculated.

[0091] It should be noted that smaller particles have a larger specific surface area and have more opportunities to contact other particles or mixer surfaces during the mixing process, thereby increasing friction. This friction is one of the main reasons for the generation of static electricity. Therefore, the smaller the equivalent diameter of the particle, the more static electricity is generated. The shape of the particles will affect the way they contact during the mixing process. More regular particles (i.e., highly round particles) usually contact and move in a more stable manner, while irregularly shaped particles may produce more friction and relative sliding, leading to greater static electricity accumulation. Therefore, in this embodiment, the particle size characteristic refers to the average particle size of the powder particles, and the shape characteristic refers to the average roundness of the particles. Then, when the particle size characteristic and the shape characteristic are larger, the less static electricity is generated during subsequent mixing.

[0092] Among them, the real-time detection data includes but is not limited to the contact potential difference value, the leakage current value and the electric field strength value. The contact potential difference value refers to the potential difference value between the initial mixture and the mixer, the leakage current value refers to the average value of the current flowing through the mixer grounding system, and the electric field strength value refers to the electric field strength value inside the mixer. The above-mentioned contact potential difference value can be directly measured by a surface potential sensor, and the leakage current value and the electric field strength value can be obtained by corresponding sensors. This embodiment does not go into too much detail on this.

[0093] The method for calculating the mixed influence value based on real-time detection data includes:

[0094]

[0095] Where MIV is the mixed influence value, Lcv is the leakage current value, Efs is the electric field strength value, Cpd is the contact potential difference value, log2[·] is the logarithmic function with base 2, sin -1 (·) is the inverse sine function, cosh(·) is the hyperbolic cosine function, exp(·) is the logarithmic function with e as the base, cot -1 [·] is the inverse tangent function, and e is a natural constant.

[0096] In this embodiment, the contact potential difference value and the electric field strength value are used as examples to illustrate that the contact potential difference value reflects the static electricity accumulation between the initial mixture and the mixer. The larger the contact potential difference value, the more serious the static electricity accumulation in the initial mixture. The electric field strength value reflects the distribution of static charge in the internal space of the mixer. The larger the electric field strength value, the more static electricity accumulation. Therefore, it can be seen from the above content that in this embodiment, the mixing influence value is positively correlated with the static electricity accumulation degree, that is, the larger the mixing influence value, the more serious the static electricity accumulation degree.

[0097] It should be added that during the mixing process, adding antistatic agents is an important means to prevent static electricity accumulation. Therefore, the mixing influence value provides a basis for calculating and determining the optimal amount of antistatic agent added. Different powder proportions, particle characteristics and mixing conditions will affect the generation and accumulation of static electricity. The mixing influence value provides real-time feedback for dynamically adjusting these parameters.

[0098] S30: inputting the proportional relationship, particle characteristics and mixing influence value into a pre-built dose generation model to obtain an antistatic dose, and determining a mixing time interval and a mixing speed interval according to the proportional relationship and the antistatic dose;

[0099] The methods for constructing the dose generation model include:

[0100] A second sample data set is obtained, wherein the second sample data set includes historical proportion relationships, historical particle characteristics, historical mixing influence values, and historical antistatic doses. The second sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical proportion relationships, historical particle characteristics, and historical mixing influence values ​​in the sample training set are used as input data of the regression network, and the historical antistatic doses in the sample training set are used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting real-time antistatic doses. The initial regression network is tested using the sample test set, and the output of the initial regression network that satisfies a preset error value is used as a dose generation model. The initial regression network is preferably a deep neural network model.

[0101] It should be added that, in the present embodiment, the dose generation model is obtained through supervised learning training, and the historical antistatic dose mentioned above is obtained by a large number of experiments conducted by technicians in this field. Since the addition of antistatic agents is to reduce the generation of static electricity in the glass powder during the production process, technicians in this field use the reduction of static electricity generation as the experimental goal to obtain the historical antistatic dose. It can be understood that from the above content, the larger the particle size characteristics and shape characteristics, the less static electricity is generated during subsequent mixing, and the larger the mixing influence value, the more serious the static electricity accumulation. Therefore, the dose of the antistatic agent can be adjusted according to the above-mentioned proportional relationship, particle characteristics and mixing influence value.

[0102] Methods for determining the mixing time interval based on the proportional relationship and antistatic dosage include:

[0103] Based on the proportional relationship, multiply the corresponding preset single-type mixing time, and then sum them up to obtain the first mixing time. Calculate the dose ratio based on the antistatic dose and the preset standard dose, multiply the first mixing time by the dose ratio to obtain the second mixing time, and use the first mixing time and the second mixing time as the endpoint values ​​of the mixing time interval respectively.

[0104] In this embodiment, it can be seen from the above content that due to the different types of sub-powders, the density and friction coefficient of the granular powders are also different. Powders with higher density are generally more likely to settle, so a longer mixing time is required to ensure their uniform distribution. On the contrary, powders with higher friction coefficients are more likely to generate static electricity during the mixing process, so they are not suitable for long-term mixing. Therefore, the above-mentioned single-type mixing time refers to the mixing time required for a single type of sub-powder. For example, when only silicon dioxide or magnesium oxide is present in the mixer, the required single-type mixing time is different. Since silicon dioxide has a lower density and a larger friction coefficient, the corresponding single-type mixing time is shorter. On the contrary, magnesium oxide has a higher density and a smaller friction coefficient, so the corresponding single-type mixing time is longer.

[0105] Among them, the role of the antistatic agent is to reduce the generation of static electricity during the production process. Therefore, adding an antistatic agent can increase the mixing time accordingly. The above-mentioned standard dosage is also pre-set based on expert experience and stored in the database. It should be added that the size of the above-mentioned first mixing time and the second mixing time is not fixed. When the first mixing time is greater than the second mixing time, the first mixing time is used as the right endpoint value of the mixing time interval, and the second mixing time is used as the left endpoint value of the mixing time interval to form a mixing time interval; when the first mixing time is less than the second mixing time, the first mixing time is used as the left endpoint value of the mixing time interval, and the second mixing time is used as the right endpoint value of the mixing time interval to form a mixing time interval.

[0106] Methods for determining the mixing speed range based on the proportional relationship and antistatic dosage include:

[0107] Based on the proportional relationship, multiply the corresponding preset single-type mixing speed, and then sum them to obtain the first mixing speed. Calculate the dosage ratio based on the antistatic dosage and the preset standard dosage, multiply the first mixing speed by the dosage ratio to obtain the second mixing speed. The first mixing speed and the second mixing speed are respectively used as the endpoint values ​​of the mixing speed range.

[0108] In this embodiment, the logic for obtaining the mixing speed interval is the same as the logic for obtaining the mixing time interval. The higher the speed during the mixing process, the easier it is to generate static electricity. For example, when only silicon dioxide or magnesium oxide exists in the mixer, the required single-type mixing speed is different. Since silicon dioxide has a lower density and a larger friction coefficient, the corresponding single-type mixing speed is slower. On the contrary, magnesium oxide has a higher density and a smaller friction coefficient, so the corresponding single-type mixing time is faster. This embodiment does not go into too much detail on this.

[0109] This embodiment can achieve adaptive control of antistatic dosage by considering proportional relationships, particle characteristics and mixing influence values. This method can provide the most suitable antistatic dosage for different mixture characteristics, powder types and static electricity accumulation during the mixing process. Different batches of raw materials may require different mixing times and rotation speeds due to differences in composition, particle characteristics, density, etc. These differences can be flexibly addressed through real-time calculation and adjustment rather than using fixed preset values. Dynamic determination of the mixing time and rotation speed range can avoid over-mixing or under-mixing and reduce unnecessary energy consumption.

[0110] S40: Determine an optimal mixing time and an optimal mixing speed based on the mixing time interval, the mixing speed interval and the optimization algorithm.

[0111] The optimization algorithm is an ant colony algorithm. Methods for determining the optimal mixing time and the optimal mixing speed according to the ant colony algorithm include:

[0112] S401: Preset the number of ants to be M, the pheromone intensity to be τ0, the volatility coefficient to be ρ, the importance coefficient of the pheromone to be α, and the importance coefficient of the heuristic information to be β;

[0113] It should be noted that the number of ants M refers to the number of ants participating in the search path in the ant colony algorithm, the pheromone intensity τ0 refers to the initial concentration of pheromone on the path, which affects the preference of ants when choosing a path, the volatility coefficient ρ usually takes a value between 0 and 1, and controls the volatility of pheromone. The importance coefficient of pheromone determines the degree of influence of pheromone on the decision of ants when choosing a path. A larger α means that ants rely more on the concentration of pheromone to choose a path. The importance coefficient of heuristic information determines the degree of influence of heuristic information on the decision of ants when choosing a path. A larger β means that ants rely more on heuristic information (such as the distance or cost of the path) to choose a path.

[0114] S402: discretizing the mixed time interval and the mixed speed interval into N nodes, and the ant selects a path between the N nodes;

[0115] Among them, this embodiment can discretize the mixed time interval and the mixed speed interval at equal intervals, and divide the mixed time interval and the mixed speed interval at equal intervals to obtain N nodes. The purpose of discretization in this embodiment is to convert the continuous search space into a finite, processable set of discrete nodes, so that the ant algorithm can search and select paths between these discrete nodes.

[0116] S403: Taking the minimization of the mixing influence value as the optimization goal, heuristic information is calculated for each node, and the heuristic information is the inverse value of the mixing time and the mixing speed, that is:

[0117]

[0118] Where η ij is the heuristic information of the ant from node i to node j, MIV ij It is the mixing influence value of the combination of mixing time and mixing speed selected by the ant from node i to node j.

[0119] Among them, the heuristic information η ij It reflects the "attractiveness" or "priority" of a path. The larger the value, the better the path. By taking the inverse of the mixing influence value to define the heuristic information, it can be ensured that ants are more inclined to choose paths with smaller mixing influence values, because this can reduce static electricity accumulation, thereby optimizing the final mixing effect. It can be understood that in this embodiment, the goal is to reduce the static electricity generated during the mixing process. Then, from the above content, it can be seen that the mixing influence value is positively correlated with the degree of static electricity accumulation, that is, the larger the mixing influence value, the more serious the static electricity accumulation. Therefore, this embodiment takes minimizing the mixing influence value as the optimization goal, and in this implementation, the longer the mixing time and the higher the mixing speed will lead to more serious static electricity accumulation. Therefore, this embodiment needs to select the optimal mixing time and mixing speed on the basis of uniform mixing, so as to maximize the reduction of static electricity accumulation.

[0120] S404: The ants select a path based on the pheromone concentration and heuristic information. The probability of going from node i to node j is:

[0121]

[0122] In the formula, p ij is the probability of an ant moving from node i to node j, τ ij is the pheromone strength from node i to node j, τ ik is the pheromone strength from node i to node k, η ik is the heuristic information from node i to node k, and allowed represents the set of paths that the ant can choose.

[0123] S405: Each ant selects a path and determines the mixed time and mixed speed in turn until the ant walks through all nodes and constructs a complete combined solution of the mixed time and mixed speed. The mixed influence value MIV is calculated for the combined solution constructed by each ant, and the combined solution with the smallest mixed influence value in the current iteration is recorded.

[0124] The process of constructing the solution simulates the foraging behavior of ants in a real environment. Through the combined effect of pheromones and heuristic information, the path selection is gradually optimized, allowing ants to find the optimal combination of mixing time and mixing speed.

[0125] S406: updating the pheromones on the path;

[0126] S407: Repeat S403-S406 until the maximum number of iterations is reached, then stop the iterations and output the mixing time and mixing speed corresponding to the path with the highest pheromone concentration as the optimal mixing time and the optimal mixing speed.

[0127] Methods for updating pheromones on a path include:

[0128] τ ij (t+1)=(1-ρ)×τ ij (t)+Δτ ij ;

[0129] In the formula, τ ij (t+1) is the pheromone strength from node i to node j in the t+1th iteration, τ ij (t) is the pheromone strength from node i to node j in the tth iteration, Δτ ij is the total pheromone increment left by all ants on the path from node i to node j in the tth iteration.

[0130] Among them, the volatilization of pheromones can avoid infinite accumulation of pheromones and ensure that the algorithm will not fall into the local optimal solution too early. Volatilization enables the ant algorithm to explore new paths instead of always following the historical optimal path.

[0131] Δτ ij The definitions include:

[0132]

[0133] Where M is the total number of ants, is the incremental pheromone of the f-th ant from node i to node j.

[0134] It is understandable that the pheromone increment mechanism ensures that the quality of the path can be reflected by the pheromone concentration, gradually guiding more ants to choose better paths and finally converge to the optimal solution.

[0135] The definitions include:

[0136]

[0137] Where MIV f is the mixed influence value corresponding to the f-th ant passing through node i to node j, and Q is a constant greater than 0.

[0138] In this embodiment, by discretizing the mixing time interval and the mixing speed interval, the ant algorithm can select paths between limited nodes and converge to the optimal solution more quickly. The goal of minimizing the mixing influence value guides the ants to choose the optimal path during the search process, and by minimizing the mixing influence value, the static electricity accumulation and agglomeration phenomenon generated during the mixing process can be effectively reduced, which directly affects the uniformity and quality of the mixing, ensuring the uniform distribution of the final mixture on each component.

[0139] In this embodiment, the proportional relationship between each sub-powder type in the mixed material is first obtained, and the initial stirring time is obtained according to the proportional relationship. The mixed material is mixed according to the initial stirring time to obtain an initial mixed material, and then a real-time surface image and real-time detection data of the initial mixed material are obtained, and particle characteristics are obtained according to the real-time surface image. The mixing influence value is calculated based on the real-time detection data, and the proportional relationship, particle characteristics and mixing influence value are input into a pre-constructed dose generation model to obtain an antistatic dose, and a mixing time interval and a mixing speed interval are determined according to the proportional relationship and the antistatic dose. The optimal mixing time and the optimal mixing speed are determined based on the mixing time interval, the mixing speed interval and the optimization algorithm. The optimized mixing time and mixing speed combination can minimize the generation of static electricity and avoid agglomeration or unevenness of the mixture, thereby improving the overall quality and consistency of the product.

[0140] Example 2

[0141] See also Figure 2 As shown, based on the same inventive concept, this embodiment discloses a multi-threaded processing system for intelligent optimization of the working parameters of a mixer. For details not provided in this embodiment, please refer to the description of the relevant parts in Embodiment 1. The system includes:

[0142] The first processing module is used to obtain the ratio relationship between the sub-powder types in the mixed material, input the ratio relationship into the pre-built time determination model, obtain the initial stirring time, and mix the mixed material according to the initial stirring time to obtain the initial mixed material;

[0143] In this embodiment, the material to be mixed can be glass powder, and the sub-powder types include at least silicon dioxide, aluminum oxide and magnesium oxide. The above-mentioned proportional relationship refers to the component proportional relationship between each sub-powder type. It can be understood that the proportional relationship between each sub-powder type is predetermined, and each sub-powder is passed into the mixer according to the proportional relationship to form the material to be mixed.

[0144] The second processing module is used to obtain a real-time surface image and real-time detection data of the initial mixture, obtain particle characteristics according to the real-time surface image, the particle characteristics at least include particle size characteristics and shape characteristics, and calculate the mixing influence value based on the real-time detection data;

[0145] Among them, the real-time surface image can be obtained through the camera unit in the mixer, and the camera unit is a camera installed in the mixer. The particle size feature mentioned above refers to the average particle size of the powder particles, and the shape feature refers to the average roundness of the particles. The particle roundness is usually defined as the square of the circumference of the particle divided by its area multiplied by a constant, and the constant can be 4π.

[0146] Methods for obtaining particle size characteristics based on real-time surface images include:

[0147] Any corner point of the real-time surface image is taken as the origin of the detection coordinate system, and the two edges connected by the corner point are taken as the two axes of the detection coordinate system. In the detection coordinate system, R groups of detection points are generated by the Markov random walk method. The equivalent diameters of the particles corresponding to the R group of detection points are extracted according to the edge detection algorithm, and the average particle size of the equivalent diameters of the R group of particles is calculated, and the average particle size is used as the particle size feature.

[0148] Methods for extracting particle equivalent diameters based on edge detection algorithms include:

[0149] The effective area of ​​the particles corresponding to the R group of detection points is extracted according to the edge detection algorithm, and the equivalent diameter of the particles is calculated based on the effective area of ​​the particles.

[0150] Methods for calculating the equivalent diameter of particles based on the effective area of ​​the particles include:

[0151]

[0152] Where EDP is the equivalent diameter of the particle, EAP is the effective area of ​​the particle, and π is a constant.

[0153] Similarly, methods for obtaining shape features based on real-time surface images include:

[0154] The effective circumference of the particles corresponding to the R group detection points is extracted according to the edge detection algorithm, and the particle roundness is obtained according to the effective area and the effective circumference of the particles. The average value of the roundness of the R group particles is used as the shape feature.

[0155] Methods for obtaining particle roundness based on particle effective area and particle effective circumference include:

[0156]

[0157] Where ROP is the roundness of the particle and EFP is the effective circumference of the particle.

[0158] The method for calculating the mixed influence value based on real-time detection data includes:

[0159]

[0160] Where MIV is the mixed influence value, Lcv is the leakage current value, Efs is the electric field strength value, Cpd is the contact potential difference value, log2[·] is the logarithmic function with base 2, sin -1 (·) is the inverse sine function, cosh(·) is the hyperbolic cosine function, exp(·) is the logarithmic function with e as the base, cot -1 [·] is the inverse tangent function, and e is a natural constant.

[0161] The third processing module is used to input the proportional relationship, particle characteristics and mixing influence value into the pre-built dose generation model to obtain the antistatic dose, and determine the mixing time interval and the mixing speed interval according to the proportional relationship and the antistatic dose;

[0162] Methods for determining the mixing time interval based on the proportional relationship and antistatic dosage include:

[0163] Based on the proportional relationship, multiply the corresponding preset single-type mixing time, and then sum them up to obtain the first mixing time. Calculate the dose ratio based on the antistatic dose and the preset standard dose, multiply the first mixing time by the dose ratio to obtain the second mixing time, and use the first mixing time and the second mixing time as the endpoint values ​​of the mixing time interval respectively.

[0164] Methods for determining the mixing speed range based on the proportional relationship and antistatic dosage include:

[0165] Based on the proportional relationship, multiply the corresponding preset single-type mixing speed, and then sum them to obtain the first mixing speed. Calculate the dosage ratio based on the antistatic dosage and the preset standard dosage, multiply the first mixing speed by the dosage ratio to obtain the second mixing speed. The first mixing speed and the second mixing speed are respectively used as the endpoint values ​​of the mixing speed range.

[0166] Optimization module: used to determine the optimal mixing time and optimal mixing speed based on the mixing time interval, mixing speed interval and optimization algorithm.

[0167] The optimization algorithm is the ant colony algorithm, and the optimization module includes initialization unit, discretization unit, calculation unit, path selection unit, combination unit, update unit and loop unit:

[0168] Initialization unit: used to preset the number of ants as M, the pheromone intensity as τ0, the volatility coefficient as ρ, the importance coefficient of pheromone as α, and the importance coefficient of heuristic information as β;

[0169] Discretization unit: used to discretize the mixed time interval and the mixed speed interval into N nodes, and the ants select paths between the N nodes;

[0170] Calculation unit: used to calculate heuristic information for each node with the minimization of the mixing influence value as the optimization goal, and the heuristic information is the inverse value of the mixing time and the mixing speed, that is:

[0171]

[0172] Where η ij is the heuristic information of the ant from node i to node j, MIV ij It is the mixing influence value of the combination of mixing time and mixing speed selected by the ant from node i to node j.

[0173] Path selection unit: used by ants to select paths based on pheromone concentration and heuristic information. The probability of going from node i to node j is:

[0174]

[0175] In the formula, p ij is the probability of an ant moving from node i to node j, τ ij is the pheromone strength from node i to node j, τ ik is the pheromone strength from node i to node k, η ik is the heuristic information from node i to node k, and allowed represents the set of paths that the ant can choose.

[0176] Combination unit: used for each ant to select a path and determine the mixing time and mixing speed in turn until the ant walks through all the nodes, constructing a complete combination solution of mixing time and mixing speed, calculating the mixed influence value MIV for the combined solution constructed by each ant, and recording the combined solution with the smallest mixed influence value in the current iteration;

[0177] Update unit: used to update the pheromone on the path;

[0178] Loop unit: used to repeat the above calculation unit to the update unit until the maximum number of iterations is reached, then stop the iteration, and output the mixing time and mixing speed corresponding to the path with the highest pheromone concentration as the optimal mixing time and optimal mixing speed.

[0179] Methods for updating pheromones on a path include:

[0180] τ ij (t+1)=(1-ρ)×τ ij (t)+Δτ ij ;

[0181] In the formula, τ ij (t+1) is the pheromone strength from node i to node j in the t+1th iteration, τ ij (t) is the pheromone strength from node i to node j in the tth iteration, Δτ ij is the total pheromone increment left by all ants on the path from node i to node j in the tth iteration.

[0182] Δτ ij The definitions include:

[0183]

[0184] Where M is the total number of ants, is the incremental pheromone of the f-th ant from node i to node j.

[0185] The definitions include:

[0186]

[0187] Where MIV f is the mixed influence value corresponding to the f-th ant passing through node i to node j, and Q is a constant greater than 0.

[0188] Example 3

[0189] like Figure 5 As shown, the present embodiment discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the multi-threaded processing method for intelligently optimizing the working parameters of the mixer is implemented.

[0190] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters, weights and thresholds in the formula are set by technicians in this field according to actual conditions.

[0191] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center through a wired network or a wireless network. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD) or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0192] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0193] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0194] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0195] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0196] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0197] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0198] Finally: 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 principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A multi-thread processing method for intelligent optimization of mixer working parameters, characterized in that: include: Obtaining the ratio relationship between the sub-powder types in the material to be mixed, inputting the ratio relationship into a pre-built time determination model to obtain an initial stirring time, and mixing the material to be mixed according to the initial stirring time to obtain an initial mixture; Acquire a real-time surface image and real-time detection data of the initial mixture, acquire particle characteristics according to the real-time surface image, the particle characteristics at least include particle size characteristics and shape characteristics, and calculate the mixing influence value based on the real-time detection data; Input the proportional relationship, particle characteristics and mixing influence value into the pre-built dose generation model to obtain the antistatic dose, and determine the mixing time interval and mixing speed interval according to the proportional relationship and the antistatic dose; The optimal mixing time and the optimal mixing speed are determined based on the mixing time interval, the mixing speed interval and the optimization algorithm.

2. The multi-thread processing method for intelligent optimization of mixer working parameters according to claim 1 is characterized in that: The particle size feature is the average value of the particle size of the powder particles, and the shape feature is the average roundness of the particles. The method for obtaining the particle size feature according to the real-time surface image includes: Any corner point of the real-time surface image is taken as the origin of the detection coordinate system, and the two edges connected by the corner point are taken as the two axes of the detection coordinate system. In the detection coordinate system, R groups of detection points are generated by the Markov random walk method. The equivalent diameters of the particles corresponding to the R group of detection points are extracted according to the edge detection algorithm, and the average particle size of the equivalent diameters of the R group of particles is calculated, and the average particle size is used as the particle size feature.

3. The multi-thread processing method for intelligent optimization of mixer working parameters according to claim 2 is characterized in that: The method for extracting the equivalent diameter of particles according to the edge detection algorithm comprises: The effective area of ​​the particles corresponding to the detection points in group R is extracted according to the edge detection algorithm, and the equivalent diameter of the particles is calculated according to the effective area of ​​the particles; Methods for calculating the equivalent diameter of particles based on the effective area of ​​the particles include: Where EDP is the equivalent diameter of the particle, EAP is the effective area of ​​the particle, and π is a constant.

4. The multi-thread processing method for intelligent optimization of mixer working parameters according to claim 2 is characterized in that: The method for acquiring shape features according to the real-time surface image comprises: The effective perimeter of the particles corresponding to the detection points of group R is extracted according to the edge detection algorithm, the roundness of the particles is obtained according to the effective area and the effective perimeter of the particles, and the average roundness of the particles of group R is taken as the shape feature; Methods for obtaining particle roundness based on particle effective area and particle effective circumference include: Where ROP is the roundness of the particle and EFP is the effective circumference of the particle.

5. The multi-thread processing method for intelligent optimization of mixer working parameters according to claim 1, characterized in that: The method for constructing the time determination model includes: A first sample data set is obtained, wherein the first sample data set includes a historical proportion relationship and a historical initial stirring time. The first sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical proportion relationship in the sample training set is used as input data of the regression network, and the historical initial stirring time in the sample training set is used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting the real-time initial stirring time. The initial regression network is tested using the sample test set, and the initial regression network that satisfies a preset error value is output as a time determination model.

6. The multi-thread processing method for intelligent optimization of mixer working parameters according to claim 2 is characterized in that: The real-time detection data includes a contact potential difference value, a leakage current value, and an electric field strength value, and a mixed influence value is calculated based on the contact potential difference value, the leakage current value, and the electric field strength value.

7. The multi-thread processing method for intelligent optimization of mixer working parameters according to claim 6 is characterized in that: The method for determining the mixing time interval according to the proportional relationship and the antistatic dosage includes: Based on the proportional relationship, multiply the corresponding preset single-type mixing time, and then sum them up to obtain the first mixing time. Calculate the dose ratio based on the antistatic dose and the preset standard dose, multiply the first mixing time by the dose ratio to obtain the second mixing time, and use the first mixing time and the second mixing time as the endpoint values ​​of the mixing time interval respectively.

8. The multi-thread processing method for intelligent optimization of mixer working parameters according to claim 7, characterized in that: The optimization algorithm is an ant colony algorithm. The method for determining the optimal mixing time and the optimal mixing speed according to the ant colony algorithm includes: S401: Preset the number of ants to be M, the pheromone intensity to be τ0, the volatility coefficient to be ρ, the importance coefficient of the pheromone to be α, and the importance coefficient of the heuristic information to be β; S402: discretizing the mixed time interval and the mixed speed interval into N nodes, and the ant selects a path between the N nodes; S403: Taking the minimization of the mixing influence value as the optimization goal, heuristic information is calculated for each node, and the heuristic information is the inverse value of the mixing time and the mixing speed, that is: Where η ij is the heuristic information of the ant from node i to node j, MIV ij The mixed influence value of the combination of mixing time and mixing speed selected by the ant from node i to node j; S404: The ants select a path based on the pheromone concentration and heuristic information. The probability of going from node i to node j is: In the formula, p ij is the probability of an ant moving from node i to node j, τ ij is the pheromone strength from node i to node j, τ ik is the pheromone strength from node i to node k, η ik is the heuristic information from node i to node k, and allowed represents the set of paths that the ant can choose; S405: Each ant selects a path and determines the mixed time and mixed speed in turn until the ant walks through all nodes and constructs a complete combined solution of the mixed time and mixed speed. The mixed influence value MIV is calculated for the combined solution constructed by each ant, and the combined solution with the smallest mixed influence value in the current iteration is recorded. S406: updating the pheromones on the path; S407: Repeat S403-S406 until the maximum number of iterations is reached, then stop the iterations and output the mixing time and mixing speed corresponding to the path with the highest pheromone concentration as the optimal mixing time and the optimal mixing speed.

9. The multi-thread processing method for intelligent optimization of mixer working parameters according to claim 8, characterized in that: The method for updating pheromones on the path includes: t ij (t+1)=(1-ρ)×τ ij (t)+Δτ ij ; In the formula, τ ij (t+1) is the pheromone strength from node i to node j in the T+1th iteration, τ ij (t) is the pheromone strength from node i to node j in the tth iteration, Δτ ij is the total pheromone increment left by all ants on the path from node i to node j in the tth iteration; Δτ ij The definitions include: Where M is the total number of ants, is the incremental pheromone of the f-th ant from node i to node j; The definitions include: Where, MIV ) is the mixed influence value corresponding to the f-th ant passing through node i to node j, and Q is a constant greater than 0.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed, the multi-threaded processing method for intelligent optimization of the working parameters of the mixer according to any one of claims 1 to 9 is implemented.

Images