A method for preparing polymer material for indoor noise reduction in buildings

Abstract

The present invention discloses a method for preparing a polymer material for indoor noise reduction in buildings, and relates to the technical field of preparation of organic polymer materials. The method comprises: obtaining the porosity and pore connectivity of a noise-reducing polymer foam sample by pore detection. If the porosity and pore connectivity are both within a threshold range, further pore distribution analysis is performed to obtain pore distribution coordinates, and the coordinates are processed in sections. The angle between each pore distribution path and the adjacent path is extracted, and a quaternion array of pore distribution is constructed in combination with the starting point and end point coordinates. The quaternion array is input into a refraction number prediction network to calculate the predicted value of the refraction number. When the predicted value reaches or exceeds the threshold, the preparation process of the polymer material is adjusted according to the relevant preparation process of the sample. The technical problem of lack of quantitative evaluation standards and refined process control in the preparation of existing polymer foam materials is solved, and the technical effect of improving the preparation accuracy and consistency of noise-reducing polymer foam materials is achieved.

Description

Technical Field

[0001] The present application relates to the technical field of organic polymer material preparation, and in particular to a method for preparing a polymer material for indoor noise reduction in a building. Background Art

[0002] Most of the current preparation methods for indoor noise reduction materials in buildings rely on empirical data, and the preparation process is usually recommended based on the frequency of similar products. However, this experience-driven process control method lacks quantitative evaluation criteria, resulting in poor precision in the preparation process, which in turn affects the stability and consistency of the material. Traditional methods cannot accurately control key properties such as the pore structure and pore connectivity of the material, making it difficult to meet the strict requirements of different building environments for noise reduction performance. In addition, since process optimization usually lacks scientific data support, the performance of existing preparation methods varies greatly between batches, making it difficult to maintain consistency and efficiency in large-scale production. Therefore, there is an urgent need for an integrated technical solution that can quantitatively control the preparation process and improve process accuracy and material performance.

[0003] In the current related technologies, there is a technical problem of lack of quantitative evaluation standards and refined process control in the preparation of polymer foam materials. Summary of the invention

[0004] The present application provides a method for preparing a polymer material for indoor noise reduction in buildings, and uses pore detection to obtain the porosity and pore connectivity of the noise-reducing polymer foam sample. If the porosity and pore connectivity are both within the threshold range, the pore distribution is further analyzed to obtain the pore distribution coordinates, and they are processed in sections. Next, the angle between each pore distribution path and the adjacent path is extracted, and the quaternion array of the pore distribution is constructed by combining the starting and end point coordinates. Then, these quaternion arrays are input into the refraction number prediction network to calculate the predicted value of the refraction number. When the predicted value reaches or exceeds the threshold, the preparation process of the polymer material is adjusted according to the relevant preparation process of the sample, thereby achieving the technical effect of improving the preparation accuracy and consistency of the noise-reducing polymer foam material.

[0005] The present application provides a method for preparing a polymer material for indoor noise reduction in buildings, comprising:

[0006] The noise reduction polymer foam sample is subjected to pore detection to obtain porosity and pore connectivity; when the porosity belongs to the porosity threshold interval and the pore connectivity belongs to the pore connectivity threshold interval, the pore distribution of the noise reduction polymer foam sample is analyzed to obtain pore distribution coordinates; the pore distribution coordinates are segmented to obtain the first pore distribution path until the Nth pore distribution path; the first angle between the first pore distribution path and the first neighborhood pore distribution path, and the first pore distribution path and the second neighborhood pore distribution path are extracted. The second angle of the distribution path, according to the coordinates of the starting point of the first pore segment, the coordinates of the end point of the first pore segment, the first angle and the second angle, construct the first segment pore distribution quaternion array; until the Nth segment pore distribution quaternion array is constructed; the first segment pore distribution quaternion array until the Nth segment pore distribution quaternion array is input into the refraction order prediction network to obtain the refraction order prediction value; when the refraction order prediction value is greater than or equal to the refraction order threshold value, the polymer material preparation control is performed according to the associated polymer material preparation process of the noise reduction polymer foam sample.

[0007] The present application also provides a computer-readable storage medium, comprising:

[0008] A computer program is stored thereon, and when the program is executed by a processor, a method for preparing a polymer material for indoor noise reduction in a building is realized.

[0009] The method for preparing a polymer material for indoor noise reduction in buildings proposed in this application first obtains the porosity and pore connectivity of the noise-reducing polymer foam sample through pore detection. If the porosity and pore connectivity are both within the threshold range, the pore distribution is further analyzed to obtain the pore distribution coordinates, and they are processed in sections. Next, the angle between each pore distribution path and the adjacent path is extracted, and the quaternion array of the pore distribution is constructed by combining the starting and ending point coordinates. These quaternion arrays are then input into the refraction number prediction network to calculate the predicted value of the refraction number. When the predicted value reaches or exceeds the threshold, the preparation process of the polymer material is adjusted according to the relevant preparation process of the sample, thereby achieving the technical effect of improving the preparation accuracy and consistency of the noise-reducing polymer foam material. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0011] Figure 1A schematic diagram of a process for preparing a polymer material for indoor noise reduction in a building provided in an embodiment of the present application;

[0012] Figure 2 A schematic diagram of a refraction order prediction network generation process for a method for preparing a polymer material for indoor noise reduction in a building provided in an embodiment of the present application. DETAILED DESCRIPTION

[0013] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0014] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0015] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0016] The present application provides a method for preparing a polymer material for indoor noise reduction in a building. Figure 1 As shown, the method includes:

[0017] Step S100, performing pore detection on the noise reduction polymer foam sample to obtain the porosity and pore connectivity. Specifically, the pores of noise reduction polymer foam samples are tested, and high-precision and high-sensitivity testing equipment is prepared to ensure the standardization of sample collection and processing. The porosity test can be carried out by the liquid saturation method, in which the sample is immersed in a non-reactive liquid of known density. The Archimedean principle is used to calculate the mass change before and after immersion and the liquid density, combined with the total volume of the sample (obtained by measuring the size); the gas adsorption method can also be used, using the adsorption characteristics of gas on the inner surface of the pores at low temperature, measuring the adsorption amount and combining it with theoretical models (such as the BET model) to calculate. The pore connectivity test can be carried out through microscopic imaging technology, such as scanning electron microscopy (SEM), to observe the microstructure image of the treated sample under electron beam irradiation, and use software analysis to statistically calculate the approximate values ​​of connected pores and total pores; X-ray computed tomography (CT) can also be used to scan the sample in three dimensions, and after the three-dimensional model is obtained by reconstruction algorithm, the software is used for quantitative analysis to obtain more accurate data, so as to obtain porosity and pore connectivity, providing key data support for subsequent processing and process optimization.

[0018] Step S200, when the porosity belongs to the porosity threshold interval, and the pore connectivity belongs to the pore connectivity threshold interval, the pore distribution analysis of the noise reduction polymer foam sample is performed to obtain the pore distribution coordinates. Specifically, when the porosity is in its threshold interval and the pore connectivity is in its threshold interval, the pore distribution analysis of the noise reduction polymer foam sample can be performed to obtain the coordinates by physical imaging methods (such as X-ray tomography (CT), scanning electron microscope (SEM)) or computational simulation methods. Taking X-ray CT as an example, the sample is first placed on the sample stage of the scanning equipment and the parameters are set. After the scanning is started, the X-ray source rotates and emits, and the detector receives the attenuation information of the X-ray penetrating the sample. The original data is converted into a three-dimensional image using a reconstruction algorithm, and the threshold is set by the image analysis software to identify the pores, and its position information constitutes the coordinate basis. When using a scanning electron microscope, the sample is first cut and pre-treated by gold spraying, and then placed in the sample chamber for scanning. The generated signal is converted into an image, and the sample is scanned and imaged at multiple angles and combined into a three-dimensional model by image stitching and analysis software, and the pore coordinates are determined according to the pixel position and scale. The computational simulation method is to build a model based on the known information of the sample, set the physical parameters, and use molecular dynamics or Monte Carlo methods to simulate the movement of polymer chains and pore evolution, and analyze the result data to extract the pore position coordinates.

[0019] In one possible implementation, when the porosity belongs to the porosity threshold interval, and the pore connectivity belongs to the pore connectivity threshold interval, the noise reduction polymer foam sample is subjected to pore distribution analysis to obtain pore distribution coordinates, and step S200 further includes step S210, when the porosity does not belong to the porosity threshold interval, or / and the pore connectivity does not belong to the pore connectivity threshold interval, or / and the predicted refractive index is less than the refractive index threshold, the associated polymer material preparation process is updated to obtain a polymer material preparation process update result. Specifically, when the porosity is not within the porosity threshold interval, or the pore connectivity is not within the pore connectivity threshold interval, or the predicted refractive index is less than the refractive index threshold, it means that there may be a problem with the current polymer material preparation process, and the related polymer material preparation process needs to be updated. The update is based on the precise judgment of the parameters, and the polymer material preparation process update results are obtained through specific algorithms and methods. It involves adjusting the parameters of each link in the material preparation, such as raw material ratio, processing temperature, pressure, reaction time, etc., so that the new preparation process meets the porosity, pore connectivity and refractive index requirements and improves material performance.

[0020] Step S220, executing a process loop according to the update result of the polymer material preparation process. Specifically, after obtaining the update result of the polymer material preparation process, executing a process loop according to the result, that is, using the new preparation process to prepare the polymer material again, and then re-performing porosity detection, pore connectivity detection and refraction number prediction on the new sample, and then judging whether the requirements are met according to the results, if not, continue to repeat the process update and process loop steps until the material meets the threshold standards in terms of porosity, pore connectivity and refraction number, so as to achieve continuous optimization and improvement of the polymer material preparation process, ensure that the quality and performance of the final product meet the standards, and the feedback loop mechanism can cope with the uncertainty and complexity of the preparation process and improve the process stability and reliability.

[0021] In a possible implementation, when the porosity does not belong to the porosity threshold interval, or / and the pore connectivity does not belong to the pore connectivity threshold interval, or / and the predicted refractive index is less than the refractive index threshold, the associated polymer material preparation process is updated to obtain the polymer material preparation process update result, and step S210 further includes step S211, obtaining the polymer material preparation history process, wherein the polymer material preparation history process is a polymer material preparation process set generated by several updates. Specifically, in the process of updating the polymer material preparation process, the first thing to do is to obtain the polymer material preparation history process, which is a set of polymer material preparation processes formed by several updates before, and the set records all the process paths tried from the beginning of preparation to the current stage. Each process update is to optimize the performance of the material, such as improving the noise reduction effect, adjusting the pore structure, etc., and the history process contains various parameter settings in this exploration process, including the type, amount, reaction conditions (temperature, pressure, time, etc.) of raw materials, and special treatment information during the processing process, which provides a rich data basis for subsequent process updates.

[0022] Step S212, when the number of historical processes of the polymer material preparation historical process is greater than or equal to the optimization population size threshold, population optimization is performed according to the polymer material preparation historical process to obtain the updated result of the polymer material preparation process. Specifically, when the number of collected historical processes of polymer material preparation is greater than or equal to the optimization population size threshold, a population optimization method will be used to obtain the updated result of the polymer material preparation process. Population optimization is a concept based on evolutionary algorithms. It regards each historical process as an individual in the population and analyzes the distribution of historical process individuals in the high-dimensional parameter space. The multiple parameters involved in each process constitute the dimension of this high-dimensional space. By calculating indicators such as the distance and similarity between individuals, their relationship in this space is evaluated. For example, some processes are similar in raw material ratios, but different in reaction temperatures, which reflects their characteristics in the parameter space. A certain relationship is determined, and according to the relationship and the pre-set optimization goal (such as increasing the porosity, improving the pore connectivity, making the refraction times closer to the threshold, etc.), a specific algorithm is used to search for a better combination of process parameters, which may involve selection operations (selecting historical process individuals with better performance), crossover operations (combining some parameters of different individuals) and mutation operations (making small-scale random changes to certain parameters of individuals), etc. Just like biological evolution, through multiple generations of evolution, one or more better polymer material preparation processes are eventually found. As an update result, a large amount of historical data is used to explore potential better processes, avoiding blind searches and improving update efficiency and success rate.

[0023] Step S213, when the number of historical processes of the polymer material preparation historical process is less than the optimization population size threshold, the associated polymer material preparation process is randomly updated to obtain the polymer material preparation process update result. Specifically, when the number of historical processes of polymer material preparation is less than the optimization population size threshold, due to limited data, it is impossible to effectively perform population optimization, and the associated polymer material preparation process is randomly updated to obtain the polymer material preparation process update result. The random update is not completely disordered, and it still adjusts the process parameters within a certain reasonable range. For example, for the amount of raw materials, it may be randomly selected within a certain percentage range of its normal usage (such as ±20%); for the reaction temperature, it is randomly selected within the temperature range that the material can withstand. Although the new method lacks sufficient data support for more refined optimization, it can provide new possibilities for process exploration in the initial stage of insufficient data, and accumulate more historical process data for the subsequent optimization, so that more effective population optimization can be performed when the data is sufficient. Random updates may discover some new and potential process directions, bringing new ideas for the improvement of polymer material preparation processes.

[0024] In a possible implementation, when the number of historical processes of the polymer material preparation historical process is greater than or equal to the optimization population size threshold, a population optimization is performed according to the polymer material preparation historical process to obtain the polymer material preparation process update result, and step S212 further includes step S2121, constructing a high-dimensional distribution space according to the attribute dimension of the polymer material preparation process. Specifically, first, a high-dimensional distribution space is constructed based on the attribute dimension of the polymer material preparation process. The polymer material preparation process often involves multiple attributes, such as the type, proportion, reaction temperature, reaction time, pressure conditions, catalyst type and dosage of raw materials. Each attribute can be regarded as a dimension, and a high-dimensional space is established with the attribute as the coordinate axis. The high-dimensional space provides a comprehensive mathematical model framework for the subsequent analysis and optimization of the polymer material preparation process, allowing us to quantitatively represent and analyze different preparation processes in a unified space.

[0025] Step S2122, distribute the polymer material preparation historical process in the high-dimensional distribution space to obtain the historical process particle distribution coordinates. Specifically, the polymer material preparation historical process is distributed in the constructed high-dimensional distribution space to obtain the historical process particle distribution coordinates. Each historical process has its unique position in this high-dimensional space, and this position is determined by its corresponding process parameters. For example, if the raw material ratio, reaction temperature and other parameters of a historical process are determined, then its coordinates in this high-dimensional space are also uniquely determined. The coordinate points represent the distribution of the preparation processes that have been tried in the entire process parameter space. They are the basic data for subsequent process update calculations and reflect the path and scope of past process exploration.

[0026] Step S2123, step 1: perform update parameter calculation according to the historical process particle distribution coordinates to obtain an update parameter set, wherein the update parameter set corresponds to the historical process particle distribution coordinates one by one. Specifically, the core of step 1 is to perform update parameter calculation according to the historical process particle distribution coordinates. For each coordinate of a historical process particle, a set of corresponding update parameters must be calculated. The calculation process comprehensively considers factors such as the position relationship between the particle and other particles in the high-dimensional space, the distance, and the distribution characteristics of the entire particle group. Through complex mathematical models and algorithms, such as calculating the weighted average of the distance between the particle and the surrounding particles, the direction vector, etc., to determine its update parameters. The update parameters reflect the improvement direction and magnitude of the current process in the high-dimensional space. Each historical process particle has its own unique update parameter set, which corresponds one by one, providing a basis for subsequent coordinate updates.

[0027] Step S2124, step 2: according to the update parameter set, the historical process particle distribution coordinates are updated to obtain the historical process particle distribution update coordinates. Specifically, in step 2, according to the update parameter set obtained in the previous step, the historical process particle distribution coordinates are updated. Specifically, according to the direction and amplitude indicated by the update parameter, the coordinates of each historical process particle are adjusted in the high-dimensional space. For example, if the update parameter indicates that a certain value needs to be increased in a certain dimension (such as the reaction temperature dimension), then the coordinates of the historical process particle in this dimension will increase accordingly. The update method is based on a comprehensive analysis of the entire historical process particle population, aiming to move each particle in a more optimal process direction, so that it is possible to find a more suitable polymer material preparation process.

[0028] Step S2125, looping step 1 to step 2 to meet the preset number of times, the historical process particle distribution update coordinates are set as the polymer material preparation process update result. Specifically, looping step 1 and step 2 to reach the preset number of times, through multiple cycles, the position of the historical process particles in the high-dimensional space is continuously adjusted and optimized, and gradually moves toward the possible optimal process area. When the preset number of cycles are completed, the historical process particle distribution update coordinates at this time are set as the polymer material preparation process update result. The result represents a new combination of preparation process parameters after multiple iterations and optimizations, which is expected to obtain better performance in the actual polymer material preparation process, such as more in line with the requirements of porosity, pore connectivity, and refractive index, etc., to achieve effective improvement of the polymer material preparation process.

[0029] In a possible implementation, step 1: perform update parameter calculation according to the historical process particle distribution coordinates to obtain an update parameter set, wherein the update parameter set corresponds one-to-one to the historical process particle distribution coordinates, and step S2123 further includes step S21231, extracting the first distribution coordinates of the historical process particle distribution coordinates. Specifically, in the process of calculating the update parameters of the polymer material preparation process, the first distribution coordinates must first be extracted from the numerous historical process particle distribution coordinates. The first distribution coordinates are the starting point for a series of subsequent calculations, representing the position information of a specific historical process in a high-dimensional distribution space, and the value of each dimension corresponds to the parameters of the process on a certain preparation process attribute. For example, the value on a certain dimension may represent the usage ratio of a certain raw material in the process, or the specific value of a certain reaction condition, etc.

[0030] Step S21232, sorting the historical process particle distribution coordinates based on the first distribution coordinates according to the preset neighborhood distance to obtain the first neighborhood distribution coordinates. Specifically, all historical process particle distribution coordinates are sorted based on the preset neighborhood distance, with the first distribution coordinates as the center. The preset neighborhood distance is an artificially set parameter used to measure the range of process similarity. By calculating the distance between each historical process particle distribution coordinate and the first distribution coordinate (this distance is the distance in high-dimensional space, which can be calculated by a specific distance formula, such as the extended form of the Euclidean distance formula in high-dimensional space), the historical process particle distribution coordinates within the preset neighborhood distance range are screened out. These screened coordinates are the first neighborhood distribution coordinates, which represent the positions of other processes in the high-dimensional space that are similar to the process corresponding to the first distribution coordinates to a certain extent.

[0031] Step S21233, calculate the separation parameter according to the first neighborhood distribution coordinates and the first distribution coordinates to obtain the first distribution coordinate separation parameter. Specifically, the separation parameter is calculated according to the first neighborhood distribution coordinates and the first distribution coordinates. The separation parameter reflects the tendency of the process represented by the first distribution coordinates to try to separate from the surrounding similar processes (the process represented by the first neighborhood distribution coordinates) in the high-dimensional space. During the calculation process, the direction and distance relationship between the first distribution coordinates and each first neighborhood distribution coordinate is comprehensively considered. For example, by calculating a certain weighted sum of the vector differences between them (the specific calculation method depends on the algorithm used), the first distribution coordinate separation parameter is obtained. The larger the value of the parameter, the more the process tends to develop in a direction different from the surrounding similar processes when updating, so as to avoid excessive concentration in locally similar process areas, which is conducive to exploring a wider process space.

[0032] Step S21234, calculate the teaming parameters according to the first neighborhood distribution coordinates and the first distribution coordinates to obtain the first distribution coordinate teaming parameters. Specifically, the teaming parameters are calculated according to the first neighborhood distribution coordinates and the first distribution coordinates to obtain the first distribution coordinate teaming parameters. The teaming parameters reflect the trend of the process represented by the first distribution coordinates to move together or maintain a relatively stable relationship with the surrounding similar processes. When calculating, the historical moving direction and speed information of the process represented by the first neighborhood distribution coordinates during the update process will be considered (this information can be obtained by recording the coordinate changes before and after each update). For example, by calculating a certain average value (such as a weighted average) of the update step size of the first neighborhood distribution coordinates, and taking its opposite as the teaming parameter. If the value of the teaming parameter is large and negative, it means that the process tends to follow the overall movement trend of the surrounding similar processes when updating, and maintains a certain "teaming" behavior, which is helpful to find a better process improvement direction in the local area.

[0033] Step S21235, foraging parameters are calculated according to the historical process particle distribution coordinates and the first distribution coordinates to obtain the first distribution coordinate foraging parameters. Specifically, foraging parameters are calculated according to the historical process particle distribution coordinates and the first distribution coordinates to obtain the first distribution coordinate foraging parameters. The foraging parameters represent the trend of the process represented by the first distribution coordinates moving toward the optimal process direction in the entire historical process group. The optimal process can be determined by some evaluation criteria, such as the process that reaches the optimal value according to the performance indicators of the material (such as porosity, pore connectivity, refraction times, etc.). During the calculation process, the foraging parameters are obtained by comparing the distance between the first distribution coordinates and the coordinates representing the optimal process (global optimal solution distribution coordinates) (also the distance in high-dimensional space). The smaller the value of this parameter, the closer the current process is to the optimal process, and the more it tends to move toward the optimal process during the update process to obtain better material preparation effects.

[0034] Step S21236, calculate the enemy avoidance parameter according to the historical process particle distribution coordinates and the first distribution coordinates to obtain the first distribution coordinate enemy avoidance parameter. Specifically, calculate the enemy avoidance parameter according to the historical process particle distribution coordinates and the first distribution coordinates to obtain the first distribution coordinate enemy avoidance parameter, which reflects the trend of the process represented by the first distribution coordinates away from the worst process in the entire historical process group, similar to the calculation of the foraging parameter, by determining the coordinates representing the worst process (global worst solution distribution coordinates), and then calculating the distance between the first distribution coordinates and the global worst solution distribution coordinates, to obtain the enemy avoidance parameter. The larger the value of this parameter, the more the current process tends to stay away from the worst process and avoid falling into a bad process area, thereby improving the efficiency and quality of process updates.

[0035] Step S21237, adding the first distribution coordinate separation parameter, the first distribution coordinate formation parameter, the first distribution coordinate foraging parameter and the first distribution coordinate avoidance parameter into the first distribution coordinate update parameter. Specifically, the calculated first distribution coordinate separation parameter, the first distribution coordinate formation parameter, the first distribution coordinate foraging parameter and the first distribution coordinate avoidance parameter are added together to form the first distribution coordinate update parameter, and the update parameter integrates the trend information of the process in separation, formation, foraging and avoidance, and comprehensively describes the update direction and strength of the process represented by the first distribution coordinate in the high-dimensional space.

[0036] Step S21238, adding the first distribution coordinate update parameter to the update parameter set. Specifically, the first distribution coordinate update parameter is added to the update parameter set, and the update parameter set includes the update parameters corresponding to all historical process particles, which will provide a basis for the subsequent update operation of the distribution coordinates of the historical process particles. Through these update parameters, the entire historical process particle group can be guided to evolve and update in the high-dimensional space towards a better polymer material preparation process.

[0037] In a possible implementation, the first distribution coordinate update parameter is added to the update parameter set, step S21238 further includes step S212381, and the separation parameter calculation formula is: ,in, Characterize the separation parameters, Characterize the first distribution coordinates, The jth distribution coordinate representing the first neighborhood distribution coordinate, represents the Euclidean distance between the first distribution coordinate and the jth distribution coordinate, and N represents the number of coordinates of the first neighborhood distribution coordinate. Specifically, the separation parameter is represented by ξ1, and its calculation formula is , the parameter reflects the separation trend of the process represented by the first distribution coordinate and the process represented by the surrounding first neighborhood distribution coordinate in high-dimensional space, where is the first distribution coordinate, which represents the position of a specific historical process that we are currently concerned about in the high-dimensional distribution space. is the jth distribution coordinate in the first neighborhood distribution coordinate, which means the position of the surrounding similar processes in the high-dimensional space centered on the first distribution coordinate and screened according to the preset neighborhood distance. is the Euclidean distance between the first distribution coordinate and the jth distribution coordinate. Euclidean distance is a common method for measuring the straight-line distance between two points in high-dimensional space. N represents the number of coordinates of the first neighborhood distribution coordinate, that is, the number of similar processes in the surrounding area. When calculating the separation parameter, the first distribution coordinate must be determined first. And all first neighbor distribution coordinates , for each first neighbor distribution coordinate , calculate it and the first distribution coordinates The Euclidean distance between , sum the distance values ​​and get , taking the opposite of this sum, we get the separation parameter For example, if there are 5 first-neighborhood distribution coordinates, their Euclidean distances with the first distribution coordinates are calculated respectively. Assuming that these distance values ​​are 2, 3, 4, 5, and 6 respectively, then their sum is 20, and the separation parameter =-20. The larger the value, the more the current process tends to separate from similar processes around it and explore a broader process space.

[0038] Step S212382, the teaming parameter calculation formula is: ,in, Characterize the parameters of the grouping, The jth distribution coordinate of the first neighborhood distribution coordinate represents the last cycle update step size, the initial update step size is equal to 0, and N represents the number of coordinates of the first neighborhood distribution coordinate. Specifically, the teaming parameter , which reflects the synergistic trend between the process represented by the first distribution coordinate and the surrounding similar processes in the updating process, that is, whether it tends to act in a "team". is the last cycle update step of the jth distribution coordinate of the first neighborhood distribution coordinate. The update step reflects the adjustment amplitude and direction of the process in the high-dimensional space. The initial update step is set to 0. N is also the number of coordinates of the first neighborhood distribution coordinate. When calculating the queuing parameters, we must first obtain the last cycle update step of each first neighborhood distribution coordinate. , and then sum the update step values ​​to get , divide this sum by the number of the first neighborhood distribution coordinates N, and finally take the opposite number to get the teaming parameter For example, if there are 4 first-neighborhood distribution coordinates, and their last cycle update steps are -1, 0, 1, and 2, respectively, then their sum is 2, and the average update step is 2 / 4=0.5. The teaming parameter =-0.5. If the value of the teaming parameter is large and negative, it indicates that the process is more inclined to follow the update trend of similar processes in the vicinity when updating, and maintain a relatively stable "teaming" relationship.

[0039] Step S212383, the foraging parameter calculation formula is: ,in, Characterize foraging parameters, Characterize the distribution coordinates of the global optimal solution, Characterizes the Euclidean distance between the global optimal solution and the first distribution coordinate. Specifically, the calculation formula of the foraging parameter is , which indicates the trend of the process represented by the first distribution coordinate moving toward the optimal process in the entire historical process group, where ξ3 is the foraging parameter, is the first distribution coordinate, is the global optimal solution distribution coordinate, which represents the position of the process with the best performance among all historical processes in the high-dimensional space. It is the Euclidean distance between the global optimal solution and the first distribution coordinate. When calculating the foraging parameters, it is only necessary to determine the distribution coordinate of the global optimal solution. and the first distribution coordinates , and then calculate the Euclidean distance between them. This distance value is the foraging parameter ξ3. For example, if the global optimal solution distribution coordinates are located at (10, 20, 30) in the high-dimensional space, and the first distribution coordinates are (12, 22, 32), the distance value calculated by the Euclidean distance formula is the foraging parameter ξ3. The smaller the parameter value, the closer the current process is to the optimal process, and the more it will be adjusted towards the optimal process during the update process.

[0040] Step S212384, the enemy avoidance parameter calculation formula is: ,in, Characterize the enemy avoidance parameters, Characterize the distribution coordinates of the global worst solution, Characterizes the Euclidean distance between the global worst solution and the first distribution coordinate. Specifically, the calculation formula of the avoidance parameter is: , the parameter reflects the trend of the process represented by the first distribution coordinate away from the worst process in the entire historical process group, where is the enemy avoidance parameter, is the first distribution coordinate, is the coordinate of the global worst solution distribution, that is, the position of the process with the worst performance among all historical processes in the high-dimensional space, is the Euclidean distance between the global worst solution and the first distribution coordinate. When calculating the avoidance parameters, first determine the global worst solution distribution coordinate and the first distribution coordinates , calculate the Euclidean distance between them , and then take its opposite number to get the enemy avoidance parameter ξ4. For example, if the global worst solution distribution coordinates in the high-dimensional space are (5, 10, 15), and the first distribution coordinates are (8, 12, 18), calculate the Euclidean distance between them and take the opposite number to get the enemy avoidance parameter ξ4. The larger the parameter value, the more the current process tends to stay away from the worst process, thereby avoiding falling into a poor process area.

[0041] In a possible implementation, step 2: according to the update parameter set, the historical process particle distribution coordinates are updated to obtain the historical process particle distribution update coordinates, and step S2124 further includes step S21241, randomly extracting the kth particle distribution coordinates according to the historical process particle distribution coordinates. Specifically, according to the historical process particle distribution coordinates, the kth particle distribution coordinates are randomly extracted, and the random extraction process ensures that each particle in the entire historical process particle population has a chance to be selected, thereby ensuring the randomness and comprehensiveness of the update process.

[0042] Step S21242, extracting the kth particle update parameters from the update parameter set, wherein the kth particle update parameters include the kth distribution coordinate separation parameter, the kth distribution coordinate formation parameter, the kth distribution coordinate foraging parameter, and the kth distribution coordinate avoidance parameter. Specifically, extracting the kth particle update parameters from the update parameter set, the update parameters are unique to the kth particle, including the kth distribution coordinate separation parameter, the kth distribution coordinate formation parameter, the kth distribution coordinate foraging parameter, and the kth distribution coordinate avoidance parameter, and the parameters describe the trend of the process represented by the kth particle in the update process from different angles, such as the tendency to separate from the surrounding particles, the tendency to act in coordination with the surrounding particles, the tendency to approach the optimal process, and the tendency to move away from the worst process.

[0043] Step S21243, wherein the k-th distribution coordinate separation parameter has a first weight factor, the k-th distribution coordinate formation parameter has a second weight factor, the k-th distribution coordinate foraging parameter has a third weight factor, and the k-th distribution coordinate avoidance parameter has a fourth weight factor. Specifically, each parameter has its corresponding weight factor, the k-th distribution coordinate separation parameter has a first weight factor, the k-th distribution coordinate formation parameter has a second weight factor, the k-th distribution coordinate foraging parameter has a third weight factor, and the k-th distribution coordinate avoidance parameter has a fourth weight factor. The weight factors are used to measure the importance of each parameter in determining the particle update step size. For example, if the first weight factor is large, it means that in the current update strategy, the separation trend has a greater impact on the update of the k-th particle; conversely, if a certain weight factor is small, the corresponding parameter has a relatively weak impact on the update.

[0044] Step S21244, when the number of k-th neighborhood distribution coordinates of the k-th particle distribution coordinate is equal to 0, the k-th distribution coordinate foraging parameter and the third weight factor, as well as the k-th distribution coordinate avoidance parameter and the fourth weight factor are fitted according to the first step length evaluation function to obtain a first recommended step length, wherein the first step length evaluation function is: ,in, Characterizes the step size required to update at the tth iteration, Characterize the third weight, Characterize the fourth weight, Characterizes the inertia weight, Characterizes the step size of the t-1th iteration update. Specifically, when the number of kth neighborhood distribution coordinates of the kth particle distribution coordinate is equal to 0, it means that there are no similar particles around the kth particle in the current high-dimensional space. According to the first step length evaluation function, the kth distribution coordinate foraging parameter and the third weight factor, as well as the kth distribution coordinate avoidance parameter and the fourth weight factor are fitted to obtain the first recommended step size. Specifically, the first step length evaluation function is: .in, Characterizes the step size required to update at the tth iteration, and are the third and fourth weights, respectively, which are related to the foraging parameters and avoidance parameters Multiplying them together reflects the importance of these two parameters in determining the step size. is the inertia weight, which takes into account the particle's motion tendency in the previous iteration, It is Through this formula, we can comprehensively consider the tendency of the particle to approach the optimal solution and move away from the worst solution as well as the previous motion state, and get a suitable first recommended step size.

[0045] Step S21245, when the number of k-th neighborhood distribution coordinates of the k-th particle distribution coordinate is greater than 0, the k-th distribution coordinate foraging parameter and the third weight factor, as well as the k-th distribution coordinate avoidance parameter and the fourth weight factor are fitted according to the second step length evaluation function to obtain a second recommended step length, wherein the second step length evaluation function is: ,in, Characterizes the step size required to update at the tth iteration, Characterize the third weight, Characterize the fourth weight, Characterize the first weight, Characterize the second weight, Characterizes the inertia weight, Characterizes the step size of the t-1th iteration update. Specifically, when the number of kth neighborhood distribution coordinates of the kth particle distribution coordinate is greater than 0, the situation is more complicated and more factors need to be considered. According to the second step size evaluation function, the kth distribution coordinate foraging parameter and the third weight factor, as well as the kth distribution coordinate avoidance parameter and the fourth weight factor are fitted to obtain the second recommended step size. The second step size evaluation function is In addition to considering the foraging parameters and the enemy avoidance parameters, the kth distribution coordinate separation parameter (weight is ) and the kth distribution coordinate grouping parameter (weight is ), because when there are neighboring particles around, the update of particles must not only consider approaching the optimal solution and moving away from the worst solution, but also consider the separation and coordination relationship with the surrounding particles. Through the formula, the second recommended step size can be determined more comprehensively based on the various trends of the particles and the previous motion state.

[0046] Step S21246, taking the global optimal solution distribution coordinates as the approach target and the global worst solution distribution coordinates as the away coordinates, randomly select the first search direction. Specifically, taking the global optimal solution distribution coordinates as the approach target and the global worst solution distribution coordinates as the away coordinates, randomly select the first search direction. This randomly selected search direction provides a direction guide for the particle update, which is determined based on a comprehensive consideration of the optimal and worst targets.

[0047] Step S21247, according to the first search direction, based on the first recommended step size or the second recommended step size, the k-th particle distribution coordinates are updated to obtain the k-th particle distribution update coordinates, and added to the historical process particle distribution update coordinates. Specifically, according to the first search direction, based on the previously obtained first recommended step size (when the number of neighborhood coordinates is 0) or the second recommended step size (when the number of neighborhood coordinates is greater than 0), the k-th particle distribution coordinates are updated to obtain the k-th particle distribution update coordinates. This updated coordinate is added to the historical process particle distribution update coordinates to complete the particle coordinate update process. The updated coordinates will participate in the next round of iterations until the preset number of iterations is met or other stop conditions are reached, thereby realizing the gradual optimization of the distribution coordinates of the entire historical process particle population to find a better polymer material preparation process.

[0048] In a possible implementation, according to the first search orientation, the k-th particle distribution coordinates are updated based on the first recommended step size or the second recommended step size to obtain the k-th particle distribution update coordinates, which are added to the historical process particle distribution update coordinates. Step S21247 further includes step S212471. When the historical process particle distribution coordinates are that the porosity does not belong to the porosity threshold interval, the fitness function is: ,in, Characterize the fitness of the particle distribution coordinates of the historical process. The greater the fitness, the better. Characterizes the smaller boundary value of the porosity threshold interval, Characterizes the larger boundary value of the porosity threshold interval, Specifically, when the porosity corresponding to the particle distribution coordinates of the historical process is not within the porosity threshold range, a specific fitness function is used to evaluate the quality of the process. The fitness function is . represents the porosity, is the smaller boundary value of the porosity threshold interval, It is the larger boundary value of the porosity threshold interval. and The porosity The boundary value of the threshold interval and The distance (the distance concept here can be a simple numerical difference), first calculate and , that is, the maximum and minimum values ​​of the distance between the porosity and the boundary value of the threshold interval. Then divide the two to get the fitness. The larger the fitness, the better the process is in terms of porosity. For example, if the porosity The maximum distance from the threshold interval boundary value is 10, and the minimum distance is 2, then the fitness The higher this value is, the greater the potential of the process in adjusting the porosity to return it to the threshold range, or the higher the degree to which the process is relatively more in line with the requirements in terms of porosity.

[0049] Step S212472, when the historical process particle distribution coordinates are that the pore connectivity does not belong to the pore connectivity threshold interval, the fitness function is: ,in, Characterize the fitness of the particle distribution coordinates of the historical process. The greater the fitness, the better. Characterizes the smaller boundary value of the pore connectivity threshold interval, Characterizes the larger boundary value of the pore connectivity threshold interval, Specifically, when the pore connectivity corresponding to the historical process particle distribution coordinates is not within the pore connectivity threshold range, the fitness function is .in, is the pore connectivity, is the smaller boundary value of the pore connectivity threshold interval, is the larger boundary value of the pore connectivity threshold interval. Similarly, and are the distances between the pore connectivity and the boundary value of the threshold interval, respectively. The calculation process is similar to the fitness function based on porosity. First, determine the maximum and minimum values ​​of the distance between the pore connectivity and the boundary value of the threshold interval, and then divide them to get the fitness The greater the fitness, the better the process performs in terms of pore connectivity. For example, pore connectivity The maximum distance from the threshold interval boundary value is 8, and the minimum distance is 4, so the fitness , the larger the value, the stronger the ability of the process in improving the pore connectivity to meet the threshold requirements.

[0050] Step S212473, when the historical process particle distribution coordinate is the refraction number prediction value which is less than the refraction number threshold, the fitness function is: ,in, Characterize the fitness of the particle distribution coordinates of the historical process. The greater the fitness, the better. Characterize the refraction number threshold, Specifically, when the predicted refraction number corresponding to the historical process particle distribution coordinates is less than the refraction number threshold, the fitness function is . represents the predicted value of the number of refractions, is the refraction number threshold, and the fitness is obtained by calculating the ratio of the predicted refraction number to the threshold. , for example, if the refraction threshold is 10, The predicted value of the number of refractions is 8, so the fitness The greater the fitness, the better. Therefore, the higher the predicted value of the number of refractions (the closer to or above the threshold), the higher the fitness of the process on this indicator, which means that the process has more advantages in meeting the requirements of the number of refractions. The fitness function provides a quantitative standard for evaluating the particle distribution coordinates of historical processes under different circumstances, which helps to determine which processes are more likely to meet the requirements of porosity, pore connectivity and number of refractions through adjustment during the process update, thereby guiding the optimization direction of the process.

[0051] Step S300, segment the pore distribution coordinates to obtain the first pore distribution path until the Nth pore distribution path. Specifically, after obtaining the pore distribution coordinates of the noise reduction polymer foam sample, it is necessary to segment it. The segmentation method should be determined based on the physical properties of the polymer material, the pore distribution law and the subsequent analysis accuracy requirements. For example, a threshold is set based on the distance between pores, and adjacent pores with a distance less than the threshold are divided into the same section, or pores showing continuous and similar geometric change trends are classified into one section based on the pore geometry and direction. Image processing technology or mathematical algorithms may be required. The first pore distribution path is determined from the starting part of the pore distribution coordinates according to the method, and the pore coordinate information and connection relationship therein are accurately recorded. Then, the coordinates are traversed backwards. When a pore that does not meet the current segment division rule is encountered, a new segment division is started. Repeat the operation until all pore distribution coordinates are divided, and the first to Nth pore distribution paths are obtained. The N value depends on the complexity of the pores and the fineness of the segmentation.

[0052] Step S400, extracting the first angle between the first pore distribution path and the first neighborhood pore distribution path, and the second angle between the first pore distribution path and the second neighborhood pore distribution path, and constructing the first pore distribution quaternion array according to the coordinates of the first pore starting point, the coordinates of the first pore ending point, the first angle and the second angle. Specifically, after completing the segmentation of the pore distribution coordinates and determining the first pore distribution path and the first and second neighborhood pore distribution paths, first clarify the representation of the pore distribution path in space, such as using vectors or converting the coordinate point sequence into vector form. Then, by calculating the dot product of the first pore distribution path vector and the neighborhood pore distribution path vector, and dividing by the product of the modulus lengths of the two vectors, the cosine value of the angle is obtained, and then the angle is obtained by the inverse trigonometric function. This method is used to obtain the first angle between the first pore distribution path and the first neighborhood pore distribution path, and the second angle between the first pore distribution path and the second neighborhood pore distribution path. Finally, the first segment pore distribution quaternion array is constructed by combining the starting point coordinates and the end point coordinates of the first segment pore. The quaternion array contains the key geometric information of the first segment pore distribution path, which is of great significance for subsequent analysis. The pore distribution quaternion arrays of other segments can be constructed in this way.

[0053] Step S500, until the Nth segment of pore distribution quaternion array is constructed. Specifically, after the pore distribution coordinate segmentation is completed, a quaternion array is constructed for each segment of the pore distribution path, from the first segment to the Nth segment. For each segment (taking the i-th segment as an example, i ranges from 1 to N), first determine its neighborhood segment, then convert it into a vector form according to the coordinate point information on the i-th segment of pore distribution path, calculate the angle between the vector and the first neighborhood segment vector (the cosine value of the angle is calculated by the vector dot product and the vector modulus relationship, and then the angle degree is obtained), and use the same method to calculate the angle with the second neighborhood segment, and then clarify the starting point and end point coordinates of the i-th segment of pores, and finally combine the starting point coordinates, the end point coordinates and the angles with the two neighborhood segments to construct the i-th segment of pore distribution quaternion array, repeat the operation until the construction of the N-th segment of pore distribution quaternion array is completed, providing pore structure information for subsequent material analysis to help study the physical properties of polymer materials.

[0054] Step S600, input the first pore distribution quaternion array to the Nth pore distribution quaternion array into the refraction number prediction network to obtain the refraction number prediction value. Specifically, a refraction number prediction network constructed and trained based on a large amount of experimental data and the known relationship between pore structure and refraction number is selected, which includes multiple neurons and different network layers, each layer is connected by a specific weight and activation function, and the mapping mode between different pore distribution characteristics and refraction number has been learned. Then, the first pore distribution quaternion array to the Nth pore distribution quaternion array is sorted to meet the network input requirements, and then these quaternion arrays are sequentially input into the refraction number prediction network according to regulations. The first layer of neurons in the network receives data and performs weighted summation and other operations, and the intermediate results generated are propagated in the network. When the data is propagated in the network, the hidden layer processes the data through complex mathematical operations and activation functions, extracts high-level features related to the refraction number, such as complex geometric relationships and distribution patterns of pores, and finally obtains the refraction number prediction value in the output layer, providing a basis for subsequent work.

[0055] In one possible implementation, Figure 2 As shown, the first pore distribution quaternion array to the Nth pore distribution quaternion array are input into the refraction number prediction network to obtain the refraction number prediction value, and step S600 further includes step S610, collecting the pore distribution quaternion array recording data and the refraction number identification data. Specifically, the pore distribution quaternion array recording data and the refraction number identification data are collected, wherein the pore distribution quaternion array recording data contains detailed information about the internal pore structure of the material, and the information may be obtained through advanced imaging technology and data analysis methods, such as using a high-resolution microscope to scan the material sample, analyze the position, size, shape and other parameters of the pores, and organize them into an array in a specific format, and the refraction number identification data is the accurate refraction number result obtained by measuring the material sample with the corresponding pore structure with professional optical measurement equipment. The quality and quantity of the data are crucial for subsequent network training, and the accuracy and representativeness of the data need to be guaranteed.

[0056] Step S620, using a mean square error loss function to evaluate the loss value of the predicted refraction order and the refraction order identification data, and inputting the pore distribution quaternion array record data into a BP neural network. Specifically, the mean square error loss function is used to evaluate the difference between the predicted refraction times and the refraction times identification data. The mean square error loss function is a commonly used method to measure the error between the predicted value and the true value. The difference is quantified by calculating the average of the square of the difference between the predicted value and the true value. Then, the collected pore distribution quaternion array record data is input into the BP neural network. The BP neural network is a multi-layer feedforward neural network based on the error back propagation algorithm. It minimizes the value of the loss function by continuously adjusting the weights and thresholds in the network. When inputting data, the data will be propagated and processed between the layers of neurons in the network. Each layer of neurons will perform a specific linear or nonlinear transformation on the input data. During the training process, the BP neural network continuously processes the input pore distribution quaternion array record data to generate a predicted refraction time, and then uses the mean square error loss function to calculate the loss value between the predicted refraction time and the refraction times identification data. The calculation is repeated. Each training will adjust the network parameters according to the loss value, continue training, and monitor the training results. When the mean square error loss value of at least 950 times in 1000 consecutive training times is less than or equal to the mean square error loss threshold, it means that the network can accurately predict the number of refractions in most cases. At this time, it can be considered that the network has achieved a good training effect.

[0057] Step S630, when the mean square error loss value is less than or equal to the mean square error loss threshold for at least 950 times of continuous 1000 trainings, the refraction number prediction network is generated. Specifically, when the above training conditions are met, the refraction number prediction network is generated, and the network has learned the intrinsic relationship between the pore distribution quaternion array record data and the refraction number identification data, and can accurately predict the refraction number based on the new pore distribution quaternion array record data, providing a reliable prediction tool for subsequent related applications, and the network can be saved and used directly when it is necessary to predict the refraction number of the material.

[0058] Step S700, when the predicted value of the refractive index is greater than or equal to the refractive index threshold, the polymer material preparation control is performed according to the associated polymer material preparation process of the noise reduction polymer foam sample. Specifically, the predicted value of the refractive index is first compared with the refractive index threshold, which is a standard value determined based on the performance requirements of the polymer material. When the predicted value is greater than or equal to the threshold, the preparation control is performed according to the associated polymer material preparation process of the noise reduction polymer foam sample. The associated polymer material preparation process includes raw material selection and ratio, parameters in the processing process (such as reaction temperature, pressure, time) and additive related information, etc. These parameters determine the material properties. During preparation, the raw materials are prepared strictly according to the formula, the reaction temperature is accurately controlled near the set value through the temperature sensor and the temperature control equipment, the appropriate pressure is maintained by the pressure regulating device, the reaction time is monitored by the timer, the additives are added according to the regulations and ensured to be evenly distributed, and the parameters of each link are continuously monitored to ensure that they are consistent with the preparation process parameter requirements, so as to obtain a material that meets the refractive index requirements.

[0059] The embodiment of the present application adopts pore detection to obtain the porosity and pore connectivity of the noise reduction polymer foam sample. If the porosity and pore connectivity are both within the threshold range, the pore distribution is further analyzed to obtain the pore distribution coordinates, and they are processed in sections. Next, the angle between each pore distribution path and the adjacent path is extracted, and the quaternion array of the pore distribution is constructed by combining the starting point and end point coordinates. Then, these quaternion arrays are input into the refraction number prediction network to calculate the predicted value of the refraction number. When the predicted value reaches or exceeds the threshold, the preparation process of the polymer material is adjusted according to the relevant preparation process of the sample, thereby achieving the technical effect of improving the preparation accuracy and consistency of the noise reduction polymer foam material.

[0060] The present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the steps of the method described in any one of the above-mentioned embodiments are implemented.

[0061] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for preparing a polymer material for indoor noise reduction in buildings, characterized in that: include: Conduct pore testing on noise reduction polymer foam samples to obtain porosity and pore connectivity; When the porosity belongs to the porosity threshold interval, and the pore connectivity belongs to the pore connectivity threshold interval, performing pore distribution analysis on the noise reduction polymer foam sample to obtain pore distribution coordinates; Segmenting the pore distribution coordinates to obtain a first pore distribution path until an Nth pore distribution path; Extracting a first angle between the first pore distribution path and the first neighborhood pore distribution path, and a second angle between the first pore distribution path and the second neighborhood pore distribution path, and constructing a first pore distribution quaternion array according to the coordinates of the first pore starting point, the first pore ending point, the first angle, and the second angle; Until the Nth segment of pore distribution quaternion array is constructed; Inputting the first segment pore distribution quaternion array to the Nth segment pore distribution quaternion array into a refraction order prediction network to obtain a refraction order prediction value; When the predicted value of the refraction number is greater than or equal to the refraction number threshold, polymer material preparation control is performed according to the polymer material preparation process associated with the noise reduction polymer foam sample.

2. The method according to claim 1, characterized in that Also includes: When the porosity does not belong to the porosity threshold interval, or / and the pore connectivity does not belong to the pore connectivity threshold interval, or / and the predicted refractive index is less than the refractive index threshold, updating the associated polymer material preparation process to obtain a polymer material preparation process update result; A process cycle is executed according to the polymer material preparation process update result.

3. The method according to claim 1, characterized in that The related polymer material preparation process is updated to obtain the polymer material preparation process update result, including: Obtaining a historical process for preparing a polymer material, wherein the historical process for preparing a polymer material is a set of polymer material preparation processes generated by a number of updates; When the number of historical processes of the polymer material preparation historical process is greater than or equal to the optimization population size threshold, a population optimization is performed according to the polymer material preparation historical process to obtain an update result of the polymer material preparation process; When the number of historical processes of the polymer material preparation historical processes is less than the optimal population size threshold, the associated polymer material preparation processes are randomly updated to obtain the polymer material preparation process update result.

4. The method according to claim 3, characterized in that When the number of historical processes of the polymer material preparation historical process is greater than or equal to the optimization population size threshold, population optimization is performed according to the polymer material preparation historical process to obtain the polymer material preparation process update result, including: Construct a high-dimensional distribution space based on the dimension of the polymer material preparation process attributes; Distributing the polymer material preparation historical process in the high-dimensional distribution space to obtain historical process particle distribution coordinates; Step 1: Calculate update parameters according to the historical process particle distribution coordinates to obtain an update parameter set, wherein the update parameter set corresponds to the historical process particle distribution coordinates one by one; Step 2: updating the historical process particle distribution coordinates according to the update parameter set to obtain the historical process particle distribution update coordinates; The steps 1 to 2 are executed repeatedly for a preset number of times, and the coordinates of the historical process particle distribution are updated and set as the updated result of the polymer material preparation process.

5. The method according to claim 4, characterized in that The updated parameters are calculated according to the historical process particle distribution coordinates to obtain an updated parameter set, including: Extracting the first distribution coordinates of the historical process particle distribution coordinates; Sorting the historical process particle distribution coordinates based on the first distribution coordinates according to a preset neighborhood distance to obtain first neighborhood distribution coordinates; Calculate a separation parameter according to the first neighborhood distribution coordinates and the first distribution coordinates to obtain a first distribution coordinate separation parameter; Calculate the teaming parameters according to the first neighborhood distribution coordinates and the first distribution coordinates to obtain the first distribution coordinate teaming parameters; Calculate foraging parameters according to the historical process particle distribution coordinates and the first distribution coordinates to obtain the first distribution coordinate foraging parameters; Calculate enemy avoidance parameters according to the historical process particle distribution coordinates and the first distribution coordinates to obtain first distribution coordinate enemy avoidance parameters; Adding the first distribution coordinate separation parameter, the first distribution coordinate teaming parameter, the first distribution coordinate foraging parameter and the first distribution coordinate avoidance parameter into the first distribution coordinate update parameter; The first distribution coordinate update parameter is added to the update parameter set.

6. The method according to claim 5, characterized in that include: The separation parameter calculation formula is: ， in, Characterize the separation parameters, Characterize the first distribution coordinates, The jth distribution coordinate representing the first neighborhood distribution coordinate, represents the Euclidean distance between the first distribution coordinate and the jth distribution coordinate, and N represents the number of coordinates of the first neighborhood distribution coordinate; The teaming parameter calculation formula is: ， in, Characterize the parameters of the grouping, The last cycle update step length of the j-th distribution coordinate representing the first neighborhood distribution coordinate, the initial update step length is equal to 0, and N represents the number of coordinates of the first neighborhood distribution coordinate; The foraging parameter calculation formula is: , in, Characterize foraging parameters, Characterize the distribution coordinates of the global optimal solution, Characterizes the Euclidean distance between the global optimal solution and the first distribution coordinate; The enemy avoidance parameter calculation formula is: ， in, Characterize the enemy avoidance parameters, Characterize the distribution coordinates of the global worst solution, Characterizes the Euclidean distance between the global worst solution and the first distribution coordinate.

7. The method according to claim 5, characterized in that The historical process particle distribution coordinates are updated according to the update parameter set to obtain the historical process particle distribution update coordinates, including: According to the historical process particle distribution coordinates, randomly extracting the kth particle distribution coordinates; Extracting a kth particle update parameter from the update parameter set, wherein the kth particle update parameter includes a kth distribution coordinate separation parameter, a kth distribution coordinate teaming parameter, a kth distribution coordinate foraging parameter, and a kth distribution coordinate avoidance parameter; wherein the kth distribution coordinate separation parameter has a first weight factor, the kth distribution coordinate formation parameter has a second weight factor, the kth distribution coordinate foraging parameter has a third weight factor, and the kth distribution coordinate avoidance parameter has a fourth weight factor; When the number of k-th neighborhood distribution coordinates of the k-th particle distribution coordinate is equal to 0, the k-th distribution coordinate foraging parameter and the third weight factor, as well as the k-th distribution coordinate avoidance parameter and the fourth weight factor are fitted according to the first step length evaluation function to obtain a first recommended step length, wherein the first step length evaluation function is: ， in, Characterizes the step size required to update at the tth iteration, Characterize the third weight, Characterize the fourth weight, Characterizes the inertia weight, Characterizes the step size of the t-1th iteration update; When the number of k-th neighborhood distribution coordinates of the k-th particle distribution coordinate is greater than 0, the k-th distribution coordinate foraging parameter and the third weight factor, as well as the k-th distribution coordinate avoidance parameter and the fourth weight factor are fitted according to the second step length evaluation function to obtain a second recommended step length, wherein the second step length evaluation function is: ， in, Characterizes the step size required to update at the tth iteration, Characterize the third weight, Characterize the fourth weight, Characterize the first weight, Characterize the second weight, Characterizes the inertia weight, Characterizes the step size of the t-1th iteration update; The global optimal solution distribution coordinates are used as the approaching target, the global worst solution distribution coordinates are used as the far-away coordinates, and the first search direction is randomly selected; According to the first search direction, the k-th particle distribution coordinates are updated based on the first recommended step size or the second recommended step size to obtain the k-th particle distribution update coordinates, which are added to the historical process particle distribution update coordinates.

8. The method according to claim 1, characterized in that The refraction number prediction network construction step comprises: Collect pore distribution quaternion array record data and refraction number identification data; The loss value of the predicted refraction number and the refraction number identification data is evaluated by a mean square error loss function, and the pore distribution quaternion array record data is input into a BP neural network; When at least 950 mean square error loss values ​​out of 1000 consecutive trainings are less than or equal to the mean square error loss threshold, the refraction number prediction network is generated.

9. The method according to claim 7, characterized in that Also includes: When the historical process particle distribution coordinates are such that the porosity does not belong to the porosity threshold interval, the fitness function is: ， in, Characterize the fitness of the particle distribution coordinates of the historical process. The greater the fitness, the better. Characterizes the smaller boundary value of the porosity threshold interval, Characterizes the larger boundary value of the porosity threshold interval, Characterize porosity; When the historical process particle distribution coordinates are such that the pore connectivity does not belong to the pore connectivity threshold interval, the fitness function is: ， in, Characterize the fitness of the particle distribution coordinates of the historical process. The greater the fitness, the better. Characterizes the smaller boundary value of the pore connectivity threshold interval, Characterizes the larger boundary value of the pore connectivity threshold interval, Characterize pore connectivity; When the historical process particle distribution coordinates are such that the predicted value of the number of refractions is less than the threshold value of the number of refractions, the fitness function is: ， in, Characterize the fitness of the particle distribution coordinates of the historical process. The greater the fitness, the better. Characterize the refraction number threshold, Characterizes the predicted value of the refraction number.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, a method for preparing a polymer material for indoor noise reduction in a building as described in any one of claims 1 to 9 is implemented.

Images