Variable working condition tool wear prediction method based on causal inference

Abstract

The invention relates to a variable-working-condition tool wear prediction method based on causal inference, and provides a method for extracting monitoring signal features based on causal inference to solve the problem that precise prediction of tool wear under variable working conditions is difficult due to the coupling action relation between machining working conditions and monitoring signals and tool wear in the numerical control machining process. The influence coefficient of abrasion and working condition change on signal change is determined by analyzing the relation among the working condition change, the signal change and the abrasion change and calculating the causal effect, partial information related to abrasion is fully extracted from signal features according to the influence coefficient, and accurate prediction of the abrasion loss of the tool is achieved. According to the method, precise prediction of variable-working-condition tool wear is realized, the influence caused by working condition changes is weakened to a great extent, and precise modeling of a tool wear prediction model is facilitated.

Description

technical field [0001] The invention relates to the field of tool wear prediction in numerical control machining, in particular to a method for predicting tool wear under variable working conditions under manufacturing big data, in particular to a method for predicting tool wear under variable working conditions based on causal inference. Background technique [0002] In the process of CNC machining, the tool is affected by multiple factors, and the wear mode is complex. The actual observable quantities are only the machining conditions, monitoring signals and tool wear, and the influence of machining conditions on tool wear and monitoring signals is coupled. The data to establish the predictive model between the monitoring signal and the tool wear is not accurate. As a measurable factor affecting tool wear, working conditions can be quantified to clarify the causal relationship between working condition changes, signal changes, and wear changes, and to find the internal cha...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the problem that there is information loss and erroneous correlation in the relationship between the decoupling working condition change, signal change and wear change by using the stability coefficient and correlation analysis in the existing method, thus affecting the prediction accuracy of the tool wear. Tool Wear Prediction Method Based on Causal Inference under Variable Working Conditions

Images