Non-uniform sample equalization method and system for product assembly process

Abstract

The invention discloses a non-uniform sample equalization method and system for a product assembly process. The method comprises the following steps: A, calculating the similarity between different samples; B, constructing a fuzzy compatibility matrix S for representing the similarity among all samples, and constructing a fuzzy compatibility space X with different grain layers through the fuzzy compatibility matrix S; C, based on a particle calculation mode, screening out a particle layer with the maximum comprehensive value of the information increment and the inter-sample similarity from thefuzzy compatible space X to serve as an optimal particle layer; D, carrying out equalization processing on the sample of the optimal particle layer. The problem of equalization of non-uniform samplesof a product assembly process is solved, and the accuracy of a final prediction result and the generalization ability of the model are improved; the problem that samples are not easy to homogenize due to strong relevance of internal data of the samples is solved, and the number of the samples in each sample particle is uniformly processed from the optimal particle layer, so the equalized effect is more representative.

Description

technical field [0001] The invention relates to the field of intelligent manufacturing, in particular to a method and system for equalizing non-uniform samples oriented to product assembly process. Background technique [0002] With the rise of intelligent manufacturing, more and more manufacturing companies apply machine learning algorithms to actual production to improve the efficiency of the production process and reduce labor costs, such as building models through machine learning during the product assembly process, To predict the production cost and time required for the product, etc. For machine learning, the initial sample data is an extremely important part. The accuracy and uniformity of the initial data will affect the accuracy and generalization ability of the final result of the machine learning algorithm. [0003] However, the assembly process of the same product of different styles has the problem of non-uniform samples. Because different styles of the same...

AI Technical Summary

Problems solved by technology

A product has a main function and multiple additional functions, and each additional function has a different prompting effect on product sales. For the additional functions that can greatly increase product sales, the manufacturer will make updates to the additional functions that have a greater sales promotion effect. The multi-style design enables more styles of products to have this additional function. For those additional functions that do not significantly increase product sales, there will be fewer product styles with this additional function, which results in uneven samples in the product assembly process. As a result, the model will be more in line with the characteristics of the assembly process of the product with a large output, and the generalization ability of the model is not enough

Moreover, for the assembly process of the same product of different styles, the strong correlation of the internal data of the sample makes it difficult to homogenize the sample

Images