Wire EDM Break Prediction With Machine Learning Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wire disconnection during electric discharge machining occurs frequently due to various machining situations, such as changes in workpiece thickness and the presence of impurities, leading to inefficient machining and reduced productivity, as existing methods require trial processes and labor-intensive adjustments.
Innovation Solution
A wire disconnection prediction device using machine learning to analyze machining data and predict disconnection risks, employing a k-nearest neighbor algorithm to classify disconnection patterns and optimize machining conditions in real-time, thereby preventing wire disconnections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional methods (trial machining, voltage threshold bypassing) are used to prevent wire disconnection, then wire disconnection can be avoided in some situations, but it requires time-consuming trial processes and labor-intensive adjustments for each workpiece type and thickness
Solution Approach 1:
The system performs preliminary learning by collecting machining data from multiple workpieces of different types and thicknesses beforehand to build a machine learning model. This preliminary action stores the knowledge of disconnection patterns and appropriate machining conditions, eliminating the need for time-consuming trial machining when processing new workpieces.
Solution Approach 2:
The invention replaces manual trial-and-error adjustment and mechanical voltage threshold monitoring with an automated machine learning system. The system uses machine learning algorithms to automatically analyze machining data, predict disconnection risks, and adjust machining conditions without human intervention, substituting manual mechanical processes with intelligent automated control.
2Reliability
If machining conditions are changed for step parts to prevent disconnection, then wire disconnection at thickness transitions can be avoided, but experiments need to be carried out beforehand for various situations where disconnection tends to occur
Solution Approach 1:
The machine learning system enables the wire electric discharge machine to automatically adjust its own machining conditions based on learned patterns from processing similar workpieces. The system self-optimizes parameters such as machining power and fluid supply according to the detected workpiece characteristics, eliminating the need for manual experimentation and setup for different situations.
Solution Approach 2:
The system dynamically changes machining parameters (power, fluid supply, pulse intervals) based on the machine learning model's predictions for each specific machining situation. Instead of using fixed parameter sets for different workpiece types, the system continuously adapts parameters according to the learned relationships between workpiece characteristics and optimal machining conditions.
3Reliability
If trial machining is performed for each plate thickness to establish appropriate machining conditions, then wire disconnection can be prevented, but it takes time and labor to cope with all kinds of situations
Solution Approach 1:
The system performs comprehensive learning experiments in advance to build a machine learning model that encompasses disconnection patterns across various workpiece types and thicknesses. This preliminary action stores the results of what would otherwise be repeated trial machining, enabling rapid prediction and adjustment during actual production without sacrificing productivity.
Solution Approach 2:
The system continuously monitors machining data and uses machine learning to predict disconnection risks in real-time. When potential disconnection is predicted, the system automatically adjusts machining conditions and provides feedback to the control system, creating a closed-loop control mechanism that prevents disconnection while maintaining high machining efficiency.
Data Source
AI summary
A wire disconnection prediction device includes: a data acquisition part configured to acquire data relating to machining of a workpiece during machining of the workpiece by a wire electric discharge machine; a preprocessing part configured to create, based on the data acquired by the data acquisition part, machining condition data, machining member data and machining state data, as state data indicating a state of the machining; and a machine learning device configured to execute, based on the state data created by the preprocessing part, processing relating to machine learning using a learning model indicating correlation between a machining state in the wire electric discharge machine and presence/absence of a possibility of disconnection occurrence of a wire electrode in the wire electric discharge machine and a disconnection cause by a plurality of class sets.