Plastic injection molding process parameter optimization method and device

Abstract

The invention discloses a plastic injection molding process parameter optimization method and device. The method comprises the following steps: constructing a fuzzy neural network; obtaining an original data set; setting optimization parameters; generating a training set according to the original data set and the optimization parameters; calculating membership layer node output of the fuzzy neural network; calculating the rule standard trigger strength of the fuzzy neural network according to the optimization parameters; training the fuzzy neural network according to the training set; and calculating the optimized output of the fuzzy neural network. According to the method, the fuzzy neural network is used for optimization and reasoning of the plastic injection molding process rule, end-to-end learning can be directly carried out in the fuzzy network, preprocessing such as data clustering or dimension reduction does not need to be carried out firstly, the problem of data redundancy in the plastic injection molding process rule optimization reasoning process is effectively solved, and the efficiency of the plastic injection molding process rule optimization reasoning is improved. The reasoning efficiency is greatly improved, and the rules in the obtained plastic injection molding rule base have high precision and good interpretability at the same time.

Description

technical field [0001] The invention belongs to the technical field of plastic injection molding, and in particular relates to a method and device for optimizing process parameters of plastic injection molding. Background technique [0002] Plastic injection molding is one of the main processing methods of plastic products, which has the advantages of short production cycle, high efficiency and easy automation. However, in actual production, injection molding is a complex multi-factor interaction process, which is affected by many factors such as raw material performance, processing equipment and environment, and involves changes in a large number of process parameters. In the production of traditional plastic injection molding process, in order to ensure the stability of product quality, the process adjustment mainly relies on the experience of craftsmen. This process is greatly affected by human factors, and it is difficult to effectively ensure production stability. [0...

AI Technical Summary

Problems solved by technology

[0004] However, the traditional expert system still has some shortcomings: on the one hand, the data volume and high dimensionality of plastic injection molding production may cause data redundancy in the intelligent network system, and there may be certain correlations between various rules , resulting in rule overfitting, making knowledge extraction difficult and inefficient; on the other hand, when existing neural networks deal with high-dimensional data, it is difficult to balance the accuracy and interpretability of rules, FCM, SSC and subtractive clustering Such clustering algorithms can significantly reduce the number of rules, but reduce the readability of the rules; data dimensionality reduction methods such as PCA and ICA can reduce the number of rules and shorten the length of the rules, but the semantic interpretability of the rules after dimensionality reduction is reduced

Images