Multi-feature thin plate part quality prediction and process parameter optimization based on machine learning

Abstract

The invention discloses a multi-quality target prediction and process parameter optimization recommendation method for numerical control machining of a multi-machining-feature thin plate part, which solves the problems of multi-quality target prediction and process parameter optimization of part machining features, and comprises the following implementation steps of collecting and sorting a data set, preprocessing the sorted data, sorting the feature importance of the data, dividing the correlation analysis data set into a training set and a test set, constructing a multi-target prediction model by using a machine learning algorithm, and constructing a technological parameter optimization model by using a genetic algorithm, and completing technological parameter optimization of the machining part. The machine learning method is used for carrying out multi-quality target prediction on the machining features, the improved genetic algorithm is used for optimizing the technological parameters, and the optimal technological parameters are obtained. The whole scheme is rigorous, complete, high in prediction precision and good in parameter optimization effect, is used for adaptive recommendation of process parameters and tool parameters, and effectively improves the machining quality.

Description

Technical field: [0001] The invention belongs to the technical field of automatic control of numerical control machining, and relates to quality prediction and parameter optimization of processed parts, in particular to quality prediction and parameter optimization of flat-panel crack antennas, and in particular to multi-quality target prediction and technology for multi-processing characteristic thin-plate parts numerical control machining The parameter optimization recommendation method is applied to the adaptive control analysis of the process parameters of parts and tool parameters in the process of CNC machining. Background technique: [0002] CNC milling is an important basic technology in intelligent manufacturing. It is widely used in automobile manufacturing, ship engineering, aircraft manufacturing and other industries. Traditional process parameter selection often relies on manual experience. Although CNC machine tools have realized the automation of the milling pr...

AI Technical Summary

Problems solved by technology

The algorithm establishes a targeted mathematical model, and adopts a reasonable optimization algorithm to achieve the best comprehensive benefit while taking into account the production cost and production time. However, when constructing a multi-objective function, the method uses empirical formulas to Express each target, and then perform linear weighted combination to form a multi-objective function. However, due to the complex processing environment and many variables involved, it will be relatively cumbersome to solve the formula for each target, and the accuracy cannot be guaranteed, which will eventually make the process parameters The reliability of the optimization results cannot meet the requirements

[0005] For the existing process parameter optimization methods for NC machining, the construction of the processing special certificate objective function mostly uses the traditional empirical formula, but in the complex machining environment and multi-parameter optimization, the expression of the empirical formula is more cumbersome and the accuracy is not high Most of the methods for optimizing process parameters use the method of orthogonal experiment, but the method of orthogonal experiment often requires more groups of experiments for the situation where the parameter variation range in actual processing is large, which will not only increase the cost, but also improve the accuracy not obvious

Images