Method for learning surface roughness of multi-parameter parts based on random forest

Abstract

The invention discloses a method for learning surface roughness of multi-parameter parts based on a random forest. Firstly, a training sample set and a test sample of a speckle image are collected, the image is preprocessed by a Butterworth filter to perform feature researches on the speckle image, multi-feature extraction is achieved, monotonicity between features and the roughness is found, anda strong classifier is constructed based on the random forest; mutual information amount learning feature weights are introduced, and combined with correlation coefficient and random forest learning parameters, a roughness learning function is constructed; finally, the learned roughness function is used for measuring the roughness of the test sample. The method establishes a novel model of a learning process type and the roughness value, breaks through the limitation that existing methods need to establish multiple measurement roughness value models for different process parts, provides a newidea for roughness measurement, and verifies effectiveness and practicability of a new algorithm through experiments.

Description

technical field [0001] The invention relates to a random forest-based multi-parameter surface roughness learning method for parts, and belongs to the technical field of surface roughness detection methods for mechanical parts. Background technique [0002] With the gradual improvement of the performance requirements for the stability, accuracy and service life of instruments and mechanical products, more and more stringent requirements are also put forward for the processing standards of the parts that constitute these products. In recent years, with the development and application of new computer-aided design (CAD) and manufacturing (CAM) and CNC machining technology in the field of mechanical production, the level of parts processing technology has reached a new height. Conventional parts size, shape and other parameters Therefore, more and more researches focus on the measurement and control of surface roughness, a microscopic shape parameter. Surface roughness has a cri...

AI Technical Summary

Problems solved by technology

However, in the existing related research, different measurement surface roughness value models are mainly established for different types of processes, and the established models only focus on the corresponding relationship between a certain speckle feature and surface roughness.

Images