An optimization method and system for material mixing

Abstract

The present invention provides an optimization method and system for material mixing, relating to the technical field of material mixing. The method includes: collecting material characteristic information; constructing an optimization space; constructing optimization conditions and an optimization function; obtaining a plurality of first design parameters and calculating to obtain a plurality of first fitness values; marking and obtaining first inferior design parameters and a plurality of first superior design parameters, deleting other design parameters with an elimination similarity greater than an elimination similarity threshold from the optimization space, and merging first superior design parameters with a merging similarity greater than a merging similarity threshold; continuing multi-round optimization of the design parameters, and after reaching preset optimization conditions, obtaining optimal design parameters for material mixing, which solves the technical problem in the prior art that due to unreasonable setting of the mixing plate, the mixing effect and mixing uniformity are poor, and achieves the technical effects of improving the optimization efficiency and accuracy, and further improving the material mixing uniformity and accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of material mixing, and specifically relates to an optimization method and system for material mixing. Background Art

[0002] The commonly used equipment for material mixing is a rotary drum. The materials move continuously in the drum as the drum rotates to complete the mixing. During the mixing process, after the materials reach a stable mixing state, the position area close to the drum wall is always occupied by small-density particles, while the large-density particles are always concentrated in the central area of the drum. There is an obvious stratification between different mixed particles. In the prior art, by adding mixing plates in the drum, the phenomenon of uneven mixing such as stratification of mixed particles is avoided. However, due to the unreasonable setting of the mixing plates, the mixing effect and the mixing uniformity are not good.

[0003] In summary, in the prior art, there is a technical problem that due to the unreasonable setting of the mixing plates, the mixing effect and the mixing uniformity are not good. Summary of the Invention

[0004] The present invention provides an optimization method and system for material mixing to solve the technical problem in the prior art that due to the unreasonable setting of the mixing plates, the mixing effect and the mixing uniformity are not good.

[0005] According to the first aspect of the present invention, an optimization method for material mixing is provided, including: collecting the material characteristic information of multiple materials to be mixed through a material information collection unit; collecting the design parameter range of the mixing plates arranged in the mixing space to construct an optimization space; constructing optimization conditions and an optimization function for optimizing the design parameters of the number, position, and height of the mixing plates through a basic condition construction unit; based on the optimization conditions and the optimization function, in the optimization space, performing an initial optimization of the design parameters through a mixing parameter optimization unit to obtain multiple first design parameters, and combining the multiple material characteristic information to perform a twin simulation of material mixing, and combining the optimization function to calculate and obtain multiple first fitness values; according to the multiple first fitness values, marking and obtaining first inferior design parameters and multiple first superior design parameters, calculating the elimination similarity between other design parameters in the optimization space and the first inferior design parameters, deleting the other design parameters with an elimination similarity greater than the elimination similarity threshold from the optimization space, calculating the merging similarity of the multiple first superior design parameters, and merging the first superior design parameters with a merging similarity greater than the merging similarity threshold to obtain a first screening result; based on the first screening result, continuing to perform multiple rounds of optimization of the design parameters, and after reaching the preset optimization conditions, obtaining the optimal design parameters, setting the mixing plates in the mixing space, and performing material mixing.

[0006] According to the second aspect of the present invention, an optimized system for material mixing is provided, including: a material characteristic acquisition module, which is used to acquire the material characteristic information of multiple materials to be mixed through a material information acquisition unit; an optimization space construction module, which is used to acquire the design parameter range of the mixing plate arranged in the mixing space and construct an optimization space; a design parameter optimization analysis module, which is used to construct optimization conditions and optimization functions for optimizing the design parameters of the number, position, and height of the mixing plate through a basic condition construction unit; a first fitness acquisition module, which is used to perform initial optimization of the design parameters in the optimization space based on the optimization conditions and optimization functions through a mixing parameter optimization unit, obtain multiple first design parameters, combine with the multiple material characteristic information, perform twin simulation of material mixing, and calculate and obtain multiple first fitness values in combination with the optimization function; a first screening result acquisition module, which is used to mark and obtain the first inferior design parameters and multiple first superior design parameters according to the multiple first fitness values, calculate the elimination similarity between other design parameters in the optimization space and the first inferior design parameters, delete the other design parameters with the elimination similarity greater than the elimination similarity threshold from the optimization space, calculate the merging similarity of the multiple first superior design parameters, and merge the first superior design parameters with the merging similarity greater than the merging similarity threshold to obtain a first screening result; an optimal design parameter acquisition module, which is used to continue to perform multiple rounds of optimization of the design parameters based on the first screening result, obtain the optimal design parameters after reaching the preset optimization conditions, set the mixing plate in the mixing space, and perform material mixing.

[0007] Based on the above analysis adopted by the present invention, the beneficial effects that can be achieved by one or more technical solutions provided by the present invention are as follows:

[0008] Collect the material characteristic information of multiple materials to be mixed, collect the design parameter range of the mixing plate set in the mixing space, construct an optimization space, and construct optimization conditions and an optimization function for optimizing the design parameters of the number, position, and height of the mixing plate. Further, based on the optimization conditions and the optimization function, in the optimization space, perform the initial optimization of the design parameters, obtain multiple first design parameters, and in combination with the multiple material characteristic information, perform twin simulation of material mixing. Combine the optimization function, calculate and obtain multiple first fitness values, mark and obtain the first inferior design parameters and multiple first superior design parameters according to the multiple first fitness values, calculate the elimination similarity between other design parameters in the optimization space and the first inferior design parameters, delete the other design parameters with the elimination similarity greater than the elimination similarity threshold from the optimization space, calculate the merging similarity of the multiple first superior design parameters, merge the first superior design parameters with the merging similarity greater than the merging similarity threshold to obtain the first screening result, and continue to perform multiple rounds of optimization of the design parameters. After reaching the preset optimization conditions, obtain the optimal design parameters and set the mixing plate in the mixing space. Thus, through the optimization of the design parameters of the mixing space of the materials, during the optimization process, by comparing the similarity between other design parameters in the optimization space and the first inferior design parameters, the design parameters with a high similarity to the first inferior design parameters are also considered as inferior design parameters, and thus they are deleted from the optimization space. Further, by performing the merging similarity comparison, the first superior design parameters with a high fitness similarity are merged and then the fitness and the optimization step size are adjusted, so as to achieve the technical effects of improving the optimization efficiency and accuracy, and further improving the material mixing uniformity and accuracy. Description of the Drawings

[0009] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. The drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0010] Figure 1 It is a schematic flowchart of an optimization method for material mixing provided by an embodiment of the present invention;

[0011] Figure 2 It is a schematic structural diagram of an optimization system for material mixing provided by an embodiment of the present invention.

[0012] Description of the reference numerals: Material characteristic acquisition module 11, optimization space construction module 12, design parameter optimization analysis module 13, first fitness acquisition module 14, first screening result acquisition module 15, optimal design parameter acquisition module 16. Detailed Implementation Modes

[0013] To make the objectives, technical solutions, and advantages of the present invention more apparent, the exemplary embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited by the exemplary embodiments described herein.

[0014] The terms used in the specification are for describing the embodiments, rather than limiting the present invention. As used in the specification, the singular terms "a", "an", and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. When used in the specification, the terms "comprise" and / or "include" specify the presence of steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other steps, operations, elements, components, and / or their groups.

[0015] Unless otherwise defined, all terms (including technical and scientific terms) used in this specification shall have the same meaning as commonly understood by those skilled in the art to which the present invention pertains. Terms, such as those defined in a common dictionary, shall not be interpreted in an idealized or overly formal sense, unless expressly defined herein. Throughout the specification, the same reference numerals represent the same elements.

[0016] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in the present invention are all information and data that have been authorized by the user or fully authorized by all parties.

[0017] Embodiment 1

[0018] Figure 1 A method for optimizing material mixing provided in an embodiment of the present invention, the method comprising:

[0019] Collecting material characteristic information of a plurality of materials to be mixed through a material information collection unit;

[0020] An embodiment of the present invention provides a method for optimizing material mixing, the method being applied to an optimization device for material mixing. The device includes a material information collection unit, a basic condition construction unit, and a mixing parameter optimization unit. The material information collection unit is used to collect basic characteristics of the materials to be mixed, and the basic condition construction unit is used to optimize the design of the device for material mixing, that is, to optimize the design of the mixing plate. The mixing parameter optimization unit is used to optimize the design parameters of the mixing plate.

[0021] Through the material information acquisition unit, acquire the material characteristic information of multiple materials to be mixed. Simply put, it is necessary to mix multiple materials together. Preferably, the multiple materials to be mixed are binary materials, such as stones and fine sand. The material characteristic information includes radius, Poisson's ratio, elastic modulus, and density. That is to say, there are various sensors in the material information acquisition unit respectively for detecting radius, Poisson's ratio, elastic modulus, and density. Thus, the material information acquisition unit acquires the radius, Poisson's ratio, elastic modulus, and density of multiple materials.

[0022] In a preferred embodiment, it further includes:

[0023] Acquire the radius, Poisson's ratio, elastic modulus, and density of multiple materials to be mixed; aggregate the radius, Poisson's ratio, elastic modulus, and density of multiple materials as multiple material characteristic information.

[0024] Through various sensors in the material information acquisition unit respectively for detecting radius, Poisson's ratio, elastic modulus, and density, acquire the radius, Poisson's ratio, elastic modulus, and density of multiple materials. Acquire the radius, Poisson's ratio, elastic modulus, and density of multiple materials to be mixed, and aggregate the radius, Poisson's ratio, elastic modulus, and density corresponding to multiple materials together as multiple material characteristic information, providing a basis for subsequent material mixing optimization.

[0025] Acquire the design parameter range of the mixing plate set in the mixing space and construct an optimization space;

[0026] Acquire the design parameter range of the mixing plate set in the mixing space and construct an optimization space. The mixing space is the device for material mixing, such as the drum of a mixer. The mixing plate is the baffle in the drum. The design parameters of the mixing plate include the number, height, and position of the mixing plate, that is, obtain the design ranges of the number, height, and position of the mixing plate as the optimization space.

[0027] In a preferred embodiment, it further includes:

[0028] According to the number, height, and position of the mixing plate set in the mixing space, obtain the number design range, height design range, and position design range; select and combine the design parameters within the number design range, height design range, and position design range to generate an optimization space.

[0029] According to the number, height, and position of mixing plates set in the mixing space, obtain the design range of the number, height, and position. Briefly speaking, installing mixing plates on the inner wall of the mixing space can make the material mixing more uniform. Then, the number, height, and position of the mixing plates have an important impact on the material mixing effect. Specifically, those skilled in the art can determine the design range of the number, height, and position of the mixing plates according to the size of the mixing space. For example, the maximum number of mixing plates that can be set, the height limit range of the mixing plates, and the positions where the mixing plates can be set, thereby obtaining the design range of the number, height, and position. Further, randomly select and combine the design parameters within the design range of the number, height, and position to obtain multiple groups of design parameters composed of different numbers, heights, and positions of mixing plates, and form an optimization space with these multiple groups of design parameters. That is to say, multiple groups of design parameters can be used for the design of mixing plates. However, the mixing plates designed with different design parameters have different effects when mixing materials. By constructing an optimization space, it provides a data basis for the subsequent optimization of design parameters.

[0030] Through the basic condition construction unit, construct optimization conditions and an optimization function for optimizing the design parameters of the number, position, and height of the set mixing plates;

[0031] Through the basic condition construction unit, construct optimization conditions and an optimization function for optimizing the design parameters of the number, position, and height of the set mixing plates. The optimization conditions can be understood as indicators for evaluating the application effects of the design parameters of the number, position, and height of the set mixing plates, and the optimization function is a fitness function for evaluating the application effects of the design parameters according to the optimization conditions.

[0032] In a preferred embodiment, it further includes:

[0033] Obtain the mixing uniformity threshold and mixing distribution threshold after mixing multiple materials; take the mixing uniformity and mixing distribution after mixing multiple materials satisfying the mixing uniformity threshold and mixing distribution threshold as optimization conditions; construct an optimization function:

[0034] ;

[0035] Wherein, is the fitness, and are weights, and are the mixing uniformity threshold and mixing distribution threshold, is the mixing uniformity information after mixing according to the design parameters, T is the number of positions for testing the material distribution after mixing, is the material distribution information at the i-th position after mixing according to the design parameters.

[0036] Obtain the mixing uniformity threshold and the mixing distribution threshold after mixing multiple materials. The mixing uniformity threshold is the standard deviation of the material distribution at multiple positions within the qualified mixing space after the materials are expected to be mixed. The mixing distribution threshold refers to the proportion threshold for the qualified mixing of materials at multiple positions within the mixing space. For example, if the input ratio of two materials is 3:7, then it is expected that the mixing ratio of the mixed materials at each position is 3:7 after mixing. The mixing distribution threshold can be understood as the target mixing ratio of the materials. If the mixing ratio of the materials at multiple positions cannot reach the mixing distribution threshold, it indicates that the mixing effect is not good. The mixing uniformity threshold refers to the standard deviation of the mixing ratio of the mixed materials at multiple positions. The smaller the standard deviation, the more uniform the mixing. That is to say, ideally, the mixing ratio of the mixed materials at each position should be the same, indicating that the mixing uniformity is ideal. However, in actual production, it is very difficult to make the mixing ratio of the mixed materials at each position exactly the same, and a small error can be allowed, which will not affect the use of the mixed materials. Therefore, based on the actual situation, the mixing uniformity threshold can be determined by combining historical experience. Taking the mixing uniformity and mixing distribution after mixing multiple materials meeting the mixing uniformity threshold and the mixing distribution threshold as the optimization conditions, an optimization function is constructed, that is, the smaller the difference between the mixing uniformity and mixing distribution and the mixing uniformity threshold and the mixing distribution threshold, the better the mixing effect is considered. The optimization function is as follows:

[0037] ;

[0038] wherein, is the fitness,[[]]END]] and are the weights, which can be subjectively set by those skilled in the art in combination with historical experience, or the weights can be analyzed through the existing weight analysis method, such as the coefficient of variation method, by collecting the historical quality inspection records of material mixing. Weight analysis is a common technical means for those skilled in the art and will not be elaborated here. and are the mixing uniformity threshold and the mixing distribution threshold,[[]]END]] is the mixing uniformity information after mixing according to the design parameters, which needs to be obtained after testing. T is the number of positions for testing the material distribution after mixing, such as T positions on the surface, middle layer, and bottom layer around the drum. is the material distribution information at the i-th position after mixing according to the design parameters, which also needs to be obtained after testing.

[0039] Thus, the construction of the optimization function is realized, providing support for the subsequent optimization of the design parameters, thereby improving the material mixing effect.

[0040] Through the mixing parameter optimization unit, based on the optimization conditions and the optimization function, in the optimization space, perform the initial optimization of the design parameters, obtain multiple first design parameters, and combine multiple material characteristic information to perform twin simulation of material mixing. Then, combine the optimization function to calculate and obtain multiple first fitness values.

[0041] Through the mixing parameter optimization unit, based on the optimization conditions and the optimization function, in the optimization space, perform the initial optimization of the design parameters, that is, randomly select multiple groups of design parameters in the optimization space as the multiple first design parameters, and combine multiple material characteristic information to perform twin simulation of material mixing. That is, based on a machine learning model, such as a neural network model, after setting the mixing space according to the multiple first design parameters, perform the simulation of material mixing, so as to obtain the corresponding mixing uniformity information and material distribution information, and then substitute them into the optimization function to calculate and obtain multiple first fitness values.

[0042] In a preferred embodiment, it further includes:

[0043] Randomly select and obtain multiple first design parameters in the optimization space; based on the material characteristic information of multiple materials, train the hybrid twin simulator; use the hybrid twin simulator to perform twin simulation of material mixing for the multiple first design parameters respectively. When the optimization conditions are met, obtain the mixing uniformity information and material distribution information after setting and mixing the mixing space with the multiple first design parameters, and combine the optimization function to calculate and obtain multiple first fitness values.

[0044] In a preferred embodiment, it further includes:

[0045] Based on the material characteristic information of multiple materials, retrieve the material mixing data records in the mixing space, obtain the sample design parameter records and T sets of sample material distribution information; use the sample design parameter records and T sets of sample material distribution information as training data, and based on machine learning, construct T hybrid twin simulation branches and train them until convergence; integrate the T hybrid twin simulation branches to obtain the hybrid twin simulator; use the hybrid twin simulator to perform twin simulation of material mixing for the multiple first design parameters, and calculate the standard deviation of the material distribution information at T positions as the mixing uniformity information. When the optimization conditions are met, combine the optimization function to calculate and obtain multiple first fitness values.

[0046] Randomly select multiple groups of design parameters in the optimization space as multiple first design parameters. Based on the material characteristic information of multiple materials, train a hybrid twin simulator based on machine learning. Use the hybrid twin simulator to perform twin simulation of material mixing on the multiple first design parameters respectively. When the optimization conditions are met, obtain the mixing uniformity information and material distribution information after mixing with the multiple first design parameters for the mixing space setting and mixing. Combine the optimization function to calculate and obtain multiple first fitness values. The optimization conditions refer to that the mixing uniformity and mixing distribution after mixing multiple materials meet the mixing uniformity threshold and mixing distribution threshold. That is to say, through the hybrid twin simulator, the mixing uniformity and mixing distribution corresponding to the multiple first design parameters can be obtained. Only when both the mixing uniformity and mixing distribution meet the mixing uniformity threshold and mixing distribution threshold, will the mixing uniformity information and material distribution information corresponding to the first design parameter be obtained and substituted into the optimization function to calculate and obtain multiple first fitness values. It can be understood that in this process, the first design parameters whose mixing uniformity and mixing distribution do not meet the mixing uniformity threshold and mixing distribution threshold will be eliminated because the mixing uniformity and mixing distribution corresponding to this part of the design parameters do not meet the mixing qualification standard. Therefore, this part of the design parameters is directly eliminated to reduce the subsequent data operation volume and improve the optimization efficiency. That is to say, the number of multiple first design parameters and multiple first fitness values can be different, and only the design parameters that meet the optimization conditions among the multiple first design parameters are selected for fitness calculation, so as to obtain the multiple first fitness values.

[0047] Among them, the process of training the hybrid twin simulator is as follows: Based on the material feature information of multiple materials, retrieve the material mixing data records in the hybrid space. The material mixing data records can be regarded as the historical records when materials with the same material feature information as the multiple materials are mixed. The material mixing data records include sample design parameter records and T sets of sample material distribution information at T positions in the corresponding hybrid space. Use the sample design parameter records and the T sets of sample material distribution information as training data, and based on machine learning, construct T hybrid twin simulation branches. For example, use a neural network model in machine learning to construct T hybrid twin simulation branches. The T hybrid twin simulation branches are used to analyze the mixing material ratios at T positions in the hybrid space. Input each set of design parameters in the sample design parameter records into the T hybrid twin simulation branches respectively, and supervise and adjust the outputs of the T hybrid twin simulation branches through the sample mixing material ratios corresponding to each set of design parameters in the T sets of sample material distribution information, so that the output data is the same as the corresponding sample mixing material ratio. After training the data in the sample design parameter records and the T sets of sample material distribution information, perform an accuracy test to obtain T hybrid twin simulation branches with the accuracy meeting the requirements. Integrate the T hybrid twin simulation branches as the hybrid twin simulator, and through the hybrid twin simulator, the mixing material distribution information at T positions can be obtained. Thus, it provides model support for the optimization analysis of material mixing.

[0048] Using the hybrid twin simulator, perform twin simulation of material mixing on the multiple first design parameters, and the T sets of material distribution information corresponding to the multiple first design parameters can be output, and calculate the standard deviation of the T sets of material distribution information as the mixing uniformity information. When the mixing uniformity information and the T sets of material distribution information meet the optimization conditions, use the optimization function to calculate and obtain multiple first fitness values.

[0049] According to the multiple first fitness values, mark and obtain the first inferior design parameters and multiple first superior design parameters, calculate the elimination similarity between other design parameters in the optimization space and the first inferior design parameters, delete the other design parameters with the elimination similarity greater than the elimination similarity threshold from the optimization space, calculate the merging similarity of the multiple first superior design parameters, and merge the first superior design parameters with the merging similarity greater than the merging similarity threshold to obtain the first screening result;

[0050] Based on multiple first fitness values, mark to obtain a first inferior design parameter and multiple first superior design parameters. Here, the first inferior design parameter refers to the first design parameter corresponding to the minimum fitness value among the multiple first fitness values, and the first design parameters corresponding to the other first fitness values except the minimum fitness value are the multiple first superior design parameters. Calculate the elimination similarity between other design parameters (excluding the first inferior design parameter) in the optimization space and the first inferior design parameter. It should be noted that during this process, a preset number of other design parameters in the optimization space can be randomly selected for the elimination similarity calculation, or all design parameters in the optimization space can be traversed for the calculation. Among them, the traversal calculation consumes a large amount of computing resources, so it can be determined by those skilled in the art. Delete the other design parameters with an elimination similarity greater than the elimination similarity threshold from the optimization space, further calculate the combined similarity of the multiple first superior design parameters, and merge the first superior design parameters with a combined similarity greater than the combined similarity threshold to obtain a first screening result.

[0051] In a preferred embodiment, it further includes:

[0052] Mark the first design parameter corresponding to the minimum value among the multiple first fitness values as the first inferior design parameter, and mark the other multiple first design parameters as multiple first superior design parameters; in the optimization space, calculate the elimination similarity between the other design parameters and the first inferior design parameter as follows:

[0053] ;

[0054] Wherein, is the elimination similarity, 、 and are weights, is the number of mixing plates in the first inferior design parameter, is the number of mixing plates in any other design parameter, is the average height of all mixing plates in the first inferior design parameter, is the average height of all mixing plates in any other design parameter, is the distance between the position of the j-th mixing plate in any other design parameter in the mixing space and the position of the nearest mixing plate in the first inferior design parameter;

[0055] Delete the design parameters with an elimination similarity greater than the elimination similarity threshold from the optimization space.

[0056] Mark the first design parameters corresponding to the minimum values within multiple first fitness values as the first inferior design parameters, and mark the other multiple first design parameters as multiple first superior design parameters. Within the optimization space, calculate the elimination similarity between the other design parameters and the first inferior design parameter, as shown in the following formula:

[0057] ;

[0058] where, is the elimination similarity, 、 and are weights, which respectively represent the weights of the number of hybrid plates, height, and position in the similarity comparison, and can be evenly distributed, is the number of hybrid plates within the first inferior design parameter, is the number of hybrid plates within any other design parameter, is the average height of all hybrid plates within the first inferior design parameter, is the average height of all hybrid plates within any other design parameter, is the distance between the position of the j-th hybrid plate within any other design parameter in the hybrid space and the position of the nearest hybrid plate within the first inferior design parameter. According to the number of hybrid plates, height, and position within the first inferior design parameter and the number of hybrid plates, height, and position within any other design parameter, calculate the elimination similarity between the other design parameters and the first inferior design parameter through the above formula. The calculation result of the elimination similarity ranges from 0 to 1. Delete the design parameters with an elimination similarity greater than the elimination similarity threshold from the optimization space. The similarity threshold is set by those skilled in the art themselves, such as 0.9. Thus, by comparing the similarity between the other design parameters within the optimization space and the first inferior design parameter, the design parameters with a high similarity to the first inferior design parameter are also considered as inferior design parameters and thus deleted, thereby improving the optimization efficiency and accuracy.

[0059] In a preferred embodiment, it further includes:

[0060] Calculate the combined similarity of multiple first optimal design parameters based on multiple first fitness values; merge two first optimal design parameters whose combined similarity is greater than the combined similarity threshold, retain the first optimal design parameter with a larger first fitness value, and perform an adjustment calculation on the first fitness value of the retained first optimal design parameter according to a preset ratio to obtain several merged first optimal design parameters and several first fitness values; respectively, use the reciprocal of the ratio of each of the several first fitness values to the mean of the several first fitness values to perform an adjustment calculation on the preset optimization step size for searching for the optimal design parameter to obtain several first optimization step sizes; use the several first optimization step sizes to perform a search adjustment on the several first optimal design parameters to perform a second-round optimization.

[0061] Calculate the combined similarity of multiple first optimal design parameters based on multiple first fitness values. Exemplarily, calculate the ratio of the smaller first fitness value and the larger first fitness value within each pair of the multiple first fitness values as the combined similarity of the two first optimal design parameters corresponding to the two first fitness values. Merge two first optimal design parameters whose combined similarity is greater than the combined similarity threshold. The combined similarity threshold is set by those skilled in the art, such as 0.9. Then retain the first optimal design parameter with a larger first fitness value. It can be understood that if the fitness values corresponding to two first optimal design parameters are relatively similar, it means that their mixing effects are close. Therefore, only retaining the first optimal design parameter with a larger first fitness value can also achieve the optimization effect, while avoiding repeated adjustment and optimization of two first optimal design parameters with relatively similar fitness values, which can improve the optimization efficiency while ensuring the optimization effect.

[0062] Perform an adjustment calculation on the first fitness value of the retained first optimal design parameter according to a preset ratio. The preset ratio is set by those skilled in the art and refers to the adjustment ratio of the first fitness value. For example, if the preset ratio is 1.1, expand the first fitness value of the first optimal design parameter according to the preset ratio to obtain the expanded first fitness value, and select the design parameter corresponding to the expanded first fitness value in the optimization space as the first optimal design parameter as well. Thus, by calculating the combined similarity between multiple first fitness values pairwise, several merged first optimal design parameters and several first fitness values can be obtained as the first screening result.

[0063] The reciprocal of the ratio of a number of first fitness values to the mean of the number of first fitness values is respectively used to adjust and calculate a preset optimization step size for searching and optimizing design parameters. The preset optimization step size includes adjustment step sizes for the number, height, and position of the mixing plates, which can be set by those skilled in the art. The preset optimization step size can be understood as an initial optimization step size. This step is to adjust the preset optimization step size. For example, each optimization increases or decreases the number of mixing plates by 4. The preset optimization step size for the number of mixing plates is 4. Based on this, multiplying the reciprocal of the ratio of a number of first fitness values to the mean of the number of first fitness values by the preset optimization step size can obtain a number of first optimization step sizes, realizing a refined adjustment of the optimization step size, reducing the optimization step size, and improving the fineness of optimization. Further, a number of first optimization step sizes are used to search and adjust the number, height, and position of the mixing plates in a number of first optimal design parameters. The adjusted number of first optimal design parameters is used for a second-round optimization. The second-round optimization repeats the foregoing steps, that is, continues to calculate the fitness of the adjusted number of first optimal design parameters, and then performs analysis of eliminating similarity and merging similarity.

[0064] Based on the first screening result, multi-round optimization of the design parameters is continued. After reaching the preset optimization conditions, the optimal design parameters are obtained, and the mixing plates in the mixing space are set to perform material mixing.

[0065] Based on the first screening result, multi-round optimization of the design parameters is continued, which is to repeat the analysis of eliminating similarity and merging similarity, adjust the design parameters. After reaching the preset optimization conditions, the preset optimization conditions are the number of optimizations, for example, set to 50 times, then 50 rounds of optimization are required. Then, according to the optimization results, the design parameter group corresponding to the maximum fitness is used as the optimal design parameters. The mixing plates in the mixing space are set according to the number, height, and position of the mixing plates in the optimal design parameters, and then material mixing is performed, thereby realizing the optimization of material mixing and improving the uniformity of material mixing.

[0066] Based on the above analysis, one or more technical solutions provided by the present invention can achieve the following beneficial effects:

[0067] Collect the material characteristic information of multiple materials to be mixed, collect the design parameter range of the mixing plate set in the mixing space, construct an optimization space, and construct optimization conditions and an optimization function for optimizing the design parameters of the number, position, and height of the mixing plate. Further, based on the optimization conditions and the optimization function, in the optimization space, perform an initial optimization of the design parameters to obtain multiple first design parameters, and combine the multiple material characteristic information to perform a twin simulation of material mixing. Combine the optimization function to calculate and obtain multiple first fitness values. According to the multiple first fitness values, mark and obtain the first inferior design parameter and multiple first superior design parameters. Calculate the elimination similarity between other design parameters in the optimization space and the first inferior design parameter, and delete other design parameters with an elimination similarity greater than the elimination similarity threshold from the optimization space. Calculate the merging similarity of the multiple first superior design parameters, and merge the first superior design parameters with a merging similarity greater than the merging similarity threshold to obtain a first screening result. Continue to perform multiple rounds of optimization of the design parameters. After reaching the preset optimization conditions, obtain the optimal design parameters and set the mixing plate in the mixing space. Thus, through the optimization of the design parameters of the material mixing space, during the optimization process, by comparing the similarity between other design parameters in the optimization space and the first inferior design parameter, design parameters with a high similarity to the first inferior design parameter are also considered inferior design parameters, and thus are deleted from the optimization space. Further, by performing a merging similarity comparison, the first superior design parameters with a high fitness similarity are merged and then the fitness and optimization step size are adjusted, so as to achieve the technical effects of improving the optimization efficiency and accuracy, and further improving the material mixing uniformity and accuracy.

[0068] Embodiment 2

[0069] Based on the same inventive concept as the optimization method for material mixing in the foregoing embodiment, as Figure 2 shown, the present invention further provides an optimization system for material mixing, and the system includes:

[0070] A material characteristic collection module 11, which is used to collect the material characteristic information of multiple materials to be mixed through a material information collection unit;

[0071] An optimization space construction module 12, which is used to collect the design parameter range of the mixing plate set in the mixing space and construct an optimization space;

[0072] A design parameter optimization analysis module 13, which is used to construct optimization conditions and an optimization function for optimizing the design parameters of the number, position, and height of the mixing plate through a basic condition construction unit;

[0073] The first fitness acquisition module 14 is configured to perform initial optimization of design parameters within the optimization space based on the optimization conditions and the optimization function through the hybrid parameter optimization unit, obtain multiple first design parameters, and perform twin simulation material mixing in combination with multiple material characteristic information, and calculate and obtain multiple first fitness values in combination with the optimization function;

[0074] The first screening result acquisition module 15 is configured to mark and obtain the first inferior design parameters and multiple first superior design parameters according to the multiple first fitness values, calculate the elimination similarity between other design parameters in the optimization space and the first inferior design parameters, delete the other design parameters with the elimination similarity greater than the elimination similarity threshold from the optimization space, calculate the merging similarity of the multiple first superior design parameters, and merge the first superior design parameters with the merging similarity greater than the merging similarity threshold to obtain the first screening result;

[0075] The optimal design parameter acquisition module 16 is configured to continue to perform multiple rounds of optimization of the design parameters based on the first screening result, obtain the optimal design parameters after reaching the preset optimization conditions, set the mixing plate in the mixing space, and perform material mixing.

[0076] Furthermore, the material characteristic acquisition module 11 is further configured to:

[0077] Collect the radii, Poisson's ratios, elastic moduli, and densities of multiple materials to be mixed;

[0078] Aggregate the radii, Poisson's ratios, elastic moduli, and densities of the multiple materials as multiple material characteristic information.

[0079] Furthermore, the optimization space construction module 12 is further configured to:

[0080] Obtain the quantity design range, height design range, and position design range according to the quantity, height, and position of the mixing plates set in the mixing space;

[0081] Select and combine the design parameters within the quantity design range, height design range, and position design range to generate an optimization space.

[0082] Furthermore, the design parameter optimization analysis module 13 is further configured to:

[0083] Obtain the mixing uniformity threshold and mixing distribution threshold after mixing multiple materials;

[0084] Use the mixing uniformity and mixing distribution after mixing multiple materials meeting the mixing uniformity threshold and mixing distribution threshold as the optimization conditions;

[0085] Construct an optimization function:

[0086] ;

[0087] Among them, is the fitness, and are the weights, and are the mixing uniformity threshold and the mixing distribution threshold, is the mixing uniformity information after mixing according to the design parameters, T is the number of positions for testing the material distribution after mixing, is the material distribution information at the i-th position after mixing according to the design parameters.

[0088] Furthermore, the first fitness acquisition module 14 is also used for:

[0089] Randomly select and obtain multiple first design parameters within the optimization space;

[0090] Train the hybrid twin simulator based on the material characteristic information of multiple materials;

[0091] Use the hybrid twin simulator to perform twin simulation of material mixing for the multiple first design parameters respectively. When the optimization conditions are met, obtain the mixing uniformity information and material distribution information after setting and mixing the mixing space with the multiple first design parameters, and calculate and obtain multiple first fitness values in combination with the optimization function.

[0092] Furthermore, the first fitness acquisition module 14 is also used for:

[0093] Based on the material characteristic information of multiple materials, retrieve the material mixing data records in the mixing space, and obtain the sample design parameter records and T sets of sample material distribution information;

[0094] Use the sample design parameter records and the T sets of sample material distribution information as training data, and construct T hybrid twin simulation branches based on machine learning and train them until convergence;

[0095] Integrate the T hybrid twin simulation branches to obtain the hybrid twin simulator;

[0096] Use the hybrid twin simulator to perform twin simulation of material mixing for the multiple first design parameters, and calculate the standard deviation of the material distribution information at T positions as the mixing uniformity information. When the optimization conditions are met, calculate and obtain multiple first fitness values in combination with the optimization function.

[0097] Furthermore, the first screening result acquisition module 15 is also used for:

[0098] Mark the first design parameters corresponding to the minimum values within multiple first fitness values as the first inferior design parameters, and mark the other multiple first design parameters as multiple first superior design parameters;

[0099] Within the optimization space, calculate the elimination similarity between the other design parameters and the first inferior design parameter, as shown in the following formula:

[0100] ;

[0101] where is the elimination similarity, , and are weights, is the number of mixing plates within the first inferior design parameter, is the number of mixing plates within any other design parameter, is the average height of all mixing plates within the first inferior design parameter, is the average height of all mixing plates within any other design parameter, is the distance between the position of the j-th mixing plate within any other design parameter in the mixing space and the position of the nearest mixing plate within the first inferior design parameter;

[0102] Delete the design parameters with an elimination similarity greater than the elimination similarity threshold from the optimization space.

[0103] Furthermore, the first screening result acquisition module 15 is further configured to:

[0104] Calculate the combined similarity of multiple first superior design parameters according to multiple first fitness values;

[0105] Merge two first superior design parameters with a combined similarity greater than the combined similarity threshold, retain the first superior design parameter with a larger first fitness value, and perform an adjustment calculation on the first fitness value of the retained first superior design parameter according to a preset ratio to obtain several merged first superior design parameters and several first fitness values;

[0106] Respectively, use the reciprocal of the ratio of several first fitness values to the mean of several first fitness values to perform an adjustment calculation on the preset optimization step size for searching for optimized design parameters to obtain several first optimization step sizes;

[0107] Use several first optimization step sizes to perform a search adjustment on several first superior design parameters for the second round of optimization.

[0108] The specific example of an optimization method for material mixing in the foregoing Embodiment 1 is equally applicable to an optimization system for material mixing in this embodiment. Through the foregoing detailed description of the optimization method for material mixing, those skilled in the art can clearly know the optimization system for material mixing in this embodiment. Therefore, for the sake of brevity of the specification, it will not be described in detail herein.

[0109] It should be understood that various forms of the processes shown above can be used, steps can be reordered, added or deleted, as long as the desired results of the technical solutions disclosed in the present invention can be achieved, and no limitations are imposed herein.

[0110] Note that the above is only a preferred embodiment of the present invention and the technical principles applied. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, re-adjustments and substitutions can be made by those skilled in the art without departing from the protection scope of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments. Without departing from the concept of the present invention, more other equivalent embodiments can be included, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. An optimization method for material mixing, characterized in that, The method is applied to an optimization device for material mixing. The device includes a material information acquisition unit, a basic condition construction unit, and a mixing parameter optimization unit. The method includes: Collecting the material characteristic information of multiple materials to be mixed through the material information acquisition unit; Collecting the design parameter range of the mixing plate in the mixing space and constructing an optimization space; Constructing an optimization condition and an optimization function for optimizing the design parameters of the number, position, and height of the mixing plate through the basic condition construction unit; Through the mixing parameter optimization unit, based on the optimization condition and the optimization function, in the optimization space, perform an initial optimization of the design parameters to obtain multiple first design parameters, and combine the multiple material characteristic information to perform a twin simulation of material mixing. Combining the optimization function, calculate and obtain multiple first fitness values. The first design parameter is a parameter combination randomly selected from the optimization space. The twin simulation of material mixing is based on a machine learning model and performs a simulation of material mixing after setting the mixing space according to the multiple first design parameters; According to the multiple first fitness values, mark and obtain the first inferior design parameter and multiple first superior design parameters, calculate the elimination similarity between other design parameters in the optimization space and the first inferior design parameter, delete the other design parameters with an elimination similarity greater than the elimination similarity threshold from the optimization space, calculate the merging similarity of the multiple first superior design parameters, and merge the first superior design parameters with a merging similarity greater than the merging similarity threshold to obtain a first screening result. Among them, calculate the ratio of the smaller first fitness value and the larger first fitness value in every two first fitness values among the multiple first fitness values as the merging similarity of the two first superior design parameters corresponding to the two first fitness values; Based on the first screening result, continue to perform multiple rounds of optimization of the design parameters. After reaching the preset optimization condition, obtain the optimal design parameters, set the mixing plate in the mixing space, and perform material mixing; Constructing an optimization condition and an optimization function for optimizing the design parameters of the number, position, and height of the mixing plate, including: Obtaining the mixing uniformity threshold and the mixing distribution threshold after mixing multiple materials; Taking the mixing uniformity and mixing distribution after mixing multiple materials to meet the mixing uniformity threshold and the mixing distribution threshold as the optimization condition; Constructing an optimization function: ； Among them, is the fitness, and is the weight, and are the mixing uniformity threshold and the mixing distribution threshold, is the mixing uniformity information after mixing according to the design parameters, T is the number of positions for testing the material distribution after mixing, is the material distribution information at the i-th position after mixing according to the design parameters; According to the multiple first fitness values, mark and obtain the first inferior design parameter and multiple first superior design parameters, calculate the elimination similarity between other design parameters in the optimization space and the first inferior design parameter, and delete the other design parameters with an elimination similarity greater than the elimination similarity threshold from the optimization space, including: Marking the first design parameter corresponding to the minimum value among the multiple first fitness values as the first inferior design parameter, and marking the other multiple first design parameters as multiple first superior design parameters; In the optimization space, calculate the elimination similarity between other design parameters and the first inferior design parameter as follows: ； Among them, To eliminate similarity, , and are weights, is the number of mixing plates within the first inferior design parameter, is the number of mixing plates within any other design parameter, is the average height of all mixing plates within the first inferior design parameter, is the average height of all mixing plates within any other design parameter, is the distance between the position of the j-th mixing plate within any other design parameter in the mixing space and the position of the nearest mixing plate within the first inferior design parameter; Deleting the design parameters with an elimination similarity greater than the elimination similarity threshold from the optimization space.

2. The method according to claim 1, characterized in that, Collecting the material characteristic information of multiple materials to be mixed, including: Collect the radii, Poisson's ratios, elastic moduli, and densities of multiple materials to be mixed. Aggregate the radii, Poisson's ratios, elastic moduli, and densities of the multiple materials as the multiple material characteristic information.

3. The method according to claim 1, characterized in that, Collect the design parameter ranges for the mixing plates set in the mixing space and construct an optimization space, including: Obtain the quantity design range, height design range, and position design range according to the quantity, height, and position of the mixing plates set in the mixing space. Select and combine the design parameters within the quantity design range, height design range, and position design range to generate the optimization space.

4. The method according to claim 1, wherein The method includes: Randomly select and obtain multiple first design parameters within the optimization space. Train the hybrid twin simulator based on the material characteristic information of the multiple materials. Use the hybrid twin simulator to perform twin simulation of material mixing for the multiple first design parameters respectively. When the optimization conditions are met, obtain the mixing uniformity information and material distribution information after setting and mixing the mixing space with the multiple first design parameters, and calculate and obtain multiple first fitness values in combination with the optimization function.

5. The method according to claim 4, wherein Based on the optimization conditions and the optimization function, perform initial optimization of the design parameters within the optimization space, obtain multiple first design parameters, and perform twin simulation of material mixing in combination with the multiple material characteristic information, and calculate and obtain multiple first fitness values in combination with the optimization function, including: Based on the material characteristic information of the multiple materials, retrieve the material mixing data records in the mixing space, and obtain the sample design parameter records and T sets of sample material distribution information. Use the sample design parameter records and the T sets of sample material distribution information as training data, and construct T hybrid twin simulation branches based on machine learning and train them until convergence. Integrate the T hybrid twin simulation branches to obtain the hybrid twin simulator. Use the hybrid twin simulator to perform twin simulation of material mixing for the multiple first design parameters, and calculate the standard deviation of the T-position material distribution information as the mixing uniformity information. When the optimization conditions are met, calculate and obtain multiple first fitness values in combination with the optimization function.

6. The method according to claim 1, characterized in that, Calculate the combined similarity of multiple first optimal design parameters, and combine the first optimal design parameters with a combined similarity greater than the combined similarity threshold, including: Calculate the combined similarity of multiple first optimal design parameters according to the multiple first fitness values. Combine two first optimal design parameters with a combined similarity greater than the combined similarity threshold, retain the first optimal design parameter with a larger first fitness value, and adjust and calculate the first fitness value of the retained first optimal design parameter according to a preset ratio to obtain several combined first optimal design parameters and several first fitness values. Respectively use the reciprocal of the ratio of each of the several first fitness values to the mean of the several first fitness values to adjust and calculate the preset optimization step size for searching for the optimized design parameters to obtain several first optimization step sizes. Use the several first optimization step sizes to search and adjust the several first optimal design parameters for the second round of optimization.

7. An optimized system for material mixing, characterized in that, For performing the steps of the method according to any one of claims 1 to 6, the system includes: Material feature acquisition module, which is used to acquire the material feature information of multiple materials to be mixed through the material information acquisition unit; Optimization space construction module, which is used to acquire the design parameter range of the mixing plate set in the mixing space and construct the optimization space; Design parameter optimization analysis module, which is used to construct the optimization conditions and optimization functions for optimizing the design parameters of the number, position, and height of the mixing plate through the basic condition construction unit; First fitness acquisition module, which is used to perform initial optimization of the design parameters in the optimization space based on the optimization conditions and optimization functions through the mixing parameter optimization unit, obtain multiple first design parameters, and combine multiple material feature information to perform twin-simulation material mixing, and calculate and obtain multiple first fitness values in combination with the optimization function. The first design parameters are parameter combinations randomly selected from the optimization space, and the twin-simulation material mixing is a simulation of material mixing after setting the mixing space according to multiple first design parameters based on a machine learning model; First screening result acquisition module, which is used to mark and obtain the first inferior design parameters and multiple first superior design parameters according to multiple first fitness values, calculate the elimination similarity between other design parameters in the optimization space and the first inferior design parameters, delete the other design parameters with the elimination similarity greater than the elimination similarity threshold from the optimization space, calculate the merging similarity of multiple first superior design parameters, and merge the first superior design parameters with the merging similarity greater than the merging similarity threshold to obtain the first screening result. Among them, the ratio of the smaller first fitness value and the larger first fitness value in every two first fitness values among multiple first fitness values is calculated as the merging similarity of the two first superior design parameters corresponding to the two first fitness values; Optimal design parameter acquisition module, which is used to continue multi-round optimization of the design parameters based on the first screening result, obtain the optimal design parameters after reaching the preset optimization conditions, set the mixing plate in the mixing space, and perform material mixing.

Images