CN-120632659-B - Method and system for optimizing injection molding parameters of mobile phone hinge

CN120632659BCN 120632659 BCN120632659 BCN 120632659BCN-120632659-B

Abstract

The invention relates to the field of data processing, in particular to a method and a system for optimizing injection molding parameters of a mobile phone hinge, wherein the method comprises the steps of obtaining an influence factor data set of mobile phone hinge injection molding; the method comprises the steps of obtaining the number of samples obtained in advance, randomly extracting samples in an influence factor data set based on the number of the samples, constructing a random forest based on the extracted samples, calculating fitting stability, calculating proper degree according to the moderate degree of the fitting stability, adjusting the number of the samples according to the fitting stability in response to the proper degree being smaller than a preset proper degree threshold, constructing the random forest according to the samples extracted again after the adjustment of the number of the samples, and taking the random forest corresponding to the number of the samples as a target random forest in response to the proper degree not being smaller than the preset proper degree threshold so as to realize mobile phone hinge injection molding processing process parameter setting. The accuracy of the mobile phone hinge injection molding processing technological parameter setting is improved.

Inventors

TAN WEI
Ouyang Songfu
CHENG CONGKUI
ZHOU RUI

Assignees

东莞华晶粉末冶金有限公司

Dates

Publication Date: 20260508
Application Date: 20250618

Claims (6)

1. The mobile phone hinge injection molding parameter optimization method is characterized by comprising the following steps of: Acquiring an influence factor data set of mobile phone hinge injection molding, and setting processing technological parameters of each influence factor data set; Obtaining the number of samples obtained in advance; randomly extracting samples in the influence factor data set based on the sampling number, constructing a random forest based on the extracted samples, inputting each sample into each random tree to obtain an output result, and obtaining the outlier degree of each sample and the corresponding set processing technological parameters, wherein the method comprises the following steps: splicing each sample with a set processing technology parameter to obtain analysis data, obtaining the distribution density of each analysis data, dividing the distribution density of each analysis data by the distribution density average value of all neighborhood data to obtain the outlier degree of the analysis data, and recording the outlier degree of the sample and the corresponding set processing technology parameter; calculating the fitting stability, comprising: ; Wherein, the Representing the difference between the set process parameters of the ith sample and the output results of the ith sample in the jth random tree, Indicating the degree of outlier of the ith sample and set processing parameters, Indicating that the first anti-zero coefficient is preset, Representing the preset second anti-zero coefficient, Representing the output of the ith sample in the jth random tree, Representing the mean value of the output results of the ith sample in all random trees, ||represents the modular length of the vector, and M and N represent the numbers of the random trees and the samples respectively; the closer the fitting stability is to the preset central value, the greater the fitting degree is, the farther the fitting stability is from the preset central value, the smaller the fitting degree is, and the fitting degree is calculated according to the moderation of the fitting stability, and the method comprises the following steps: Normalizing the fitting stability, taking the reciprocal of the absolute value of the difference between the normalized fitting stability and a preset central value, and normalizing to obtain the proper degree of the sampling number; The method comprises the steps of multiplying a preset adjustment step length by a difference value between the preset adjustment step length and the normalized fitting stability to obtain an adjusted number, and adding the adjusted number to the sample number to obtain an adjusted sample number, and multiplying the preset adjustment step length by the normalized fitting stability to obtain an adjusted number if the normalized fitting stability is greater than the preset central value; Constructing a random forest according to the samples re-extracted by the adjusted sampling quantity; And responding to the suitability degree not smaller than a preset suitability degree threshold, and taking the random forest corresponding to the sampling number as a target random forest to realize the mobile phone hinge injection molding processing technological parameter setting.
2. The method for optimizing parameters of hinge injection molding of a mobile phone according to claim 1, wherein the obtaining the number of samples obtained in advance comprises: And setting the influence factor data with the same processing technological parameters in the influence factor data set as one type, acquiring boundary data between every two types, acquiring boundary definition of every two types and boundary rule complexity according to the boundary data, taking a quotient of the boundary rule complexity and the boundary definition as the distinguishing difficulty of every two types, taking a distinguishing difficulty mean value of all two types as the whole distinguishing difficulty, and setting the sampling number according to the whole distinguishing difficulty, wherein the sampling number is positively correlated with the whole distinguishing difficulty.
3. The method for optimizing parameters of hinge injection molding of a mobile phone according to claim 2, wherein the method for obtaining the definition of the boundary comprises the following steps: Obtaining the geometric centers of the influence factor data of each category, taking the geometric centers of the influence factor data of two categories as straight lines, projecting the influence factor data of two categories to the straight lines to obtain projection points, arranging category mark values of all the influence factor data of the two categories according to the arrangement sequence of the projection points to obtain a category mark value sequence, and marking the inverse of the arrangement entropy of the category mark value sequence as the definition of the boundary of the two categories.
4. The method for optimizing mobile phone hinge injection molding parameters according to claim 3, wherein the method for obtaining the category flag value comprises the following steps: The category flag value of the influence factor data of one category is set to 1, and the category flag value of the influence factor data of another category is set to 0.
5. The method for optimizing parameters of mobile phone hinge injection molding according to claim 2, wherein the method for obtaining complexity of the boundary rule comprises the following steps: setting a term number variable Fitting a polynomial with the number of T by using boundary points, sequentially taking different values of T according to the sequence from small to large, obtaining fitting errors of polynomials corresponding to each value of T, and obtaining a polynomial with the fitting errors smaller than a preset error threshold value for the first time as a target polynomial; And multiplying the target degree by the number of terms of the target polynomial to obtain the complexity of the boundary rule.
6. A mobile phone hinge injection molding parameter optimization system comprising a processor and a memory, the memory storing computer program instructions that when executed by the processor implement a mobile phone hinge injection molding parameter optimization method according to any one of claims 1-5.

Description

Method and system for optimizing injection molding parameters of mobile phone hinge Technical Field The present invention relates to the field of data processing. More particularly, the invention relates to a mobile phone hinge injection molding parameter optimization method and system. Background With the rapid expansion of the folding screen mobile phone market, the mobile phone hinge is taken as a core component, and the quality and the performance of the mobile phone hinge are paid attention to. The hinge structure of the mobile phone is precise and complex, and has extremely high requirements on the injection molding process. The accurate injection molding parameters are key to guaranteeing the quality of the hinge, and not only affect the dimensional accuracy and the mechanical property, but also relate to the production efficiency and the cost control. In the prior art, a method for determining the injection molding process parameters of the hinge of the mobile phone has a plurality of limitations. The traditional mode mainly depends on manual experience and repeated trial and error, and lacks of systemicity and accuracy. The process personnel preliminarily set parameters according to past experience, and then the process is complicated, time-consuming and high in cost through repeated die test adjustment. The mode is difficult to adapt to the design requirements of complicated and changeable mobile phone hinges, especially for novel materials and complex structures, the trial-and-error frequency is exponentially increased, and the efficiency is extremely low. In recent years, machine learning technology is rapidly rising, and a new thought and method are provided for solving the problems. The random forest algorithm is used as a high-efficiency machine learning algorithm, and is widely applied in various fields by virtue of the excellent performance of the random forest algorithm in the process of multi-variable complex problems. However, whether the random forest algorithm can construct a high-precision random forest model is related to the proper degree of sampling quantity setting, wherein when the sampling quantity setting is too large, the number of samples for constructing the random tree is large, and fitting is further caused, and when the sampling quantity setting is too small, the situation that the number of samples for constructing the random tree is small, and fitting capability of the random tree is insufficient is caused easily. How to set the appropriate number of samples is thus the focus of the study of the present invention. Disclosure of Invention In order to solve the problem of how to set a proper sampling number, the invention provides a mobile phone hinge injection molding parameter optimization method and a mobile phone hinge injection molding parameter optimization system. In a first aspect, the present invention provides a method for optimizing injection molding parameters of a hinge of a mobile phone, including: Acquiring an influence factor data set of mobile phone hinge injection molding, and setting processing technological parameters of each influence factor data set; randomly extracting samples in an influence factor dataset based on the sampling number, constructing a random forest based on the extracted samples, inputting each sample into each random tree to obtain an output result, obtaining each sample and the outlier degree corresponding to the set processing technological parameter, and calculating fitting stability, wherein the fitting stability is inversely related to the difference between the set processing technological parameter and the output result and is positively related to the outlier degree; And setting parameters of the mobile phone hinge injection molding process by taking the random forest corresponding to the sampling number as a target random forest in response to the suitability degree not being smaller than the preset suitability degree threshold value. The method and the device improve the fitting accuracy of the random forest model by adaptively setting the sampling number, so that more proper processing parameters are set for the mobile phone hinge, further, the fitting condition of the random forest model is analyzed to judge the proper condition of the sampling number, the sampling number is adjusted according to the proper condition of the sampling number, the accurate sampling number is obtained in a self-adaptive mode, further, the fitting proper condition of the random forest model is reflected by introducing fitting stability when the fitting condition of the random forest model is analyzed, a basis is provided for the adjustment of the sampling number, and further, the fitting condition of the random forest model is reflected by introducing outlier degree and setting the difference between the processing parameters and an output result when the fitting stability is analyzed, so that the fitting condition of the under fitting and t