Wire EDM Thermal Displacement Correction via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wire electric discharge machines face challenges in maintaining machining precision due to thermal deformation of machine elements, with current thermal displacement correction methods being either inefficient or overly complex, particularly in determining optimal sensor placement and calculating displacement amounts.
Innovation Solution
A wire electric discharge machine equipped with temperature detection and position measurement systems, using machine learning to associate temperature data with guide positions, allowing for accurate calculation and correction of thermal displacement, and optionally adjusting sensor placement based on influence analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If temperature sensors are installed to machine elements with great influence on thermal displacement, then thermal displacement correction accuracy is improved, but device complexity and cost increase
Solution Approach 1:
The patent changes the parameter of sensor installation positions by using machine learning to identify optimal locations based on thermal influence analysis. Instead of installing sensors on all machine elements or following conventional fixed positions, the system calculates thermal displacement amounts for various positions and selects optimal sensor locations that maximize correction accuracy while minimizing the number of sensors needed.
Solution Approach 2:
The machine tool itself performs the analysis to determine optimal sensor positions. The control device uses the machine's own thermal displacement data and machine learning algorithms to identify which machine elements have the greatest thermal influence, thereby self-determining the best sensor installation locations without external intervention.
2Device complexity
If temperature sensors are installed to machine elements with almost no influence on thermal displacement, then device complexity is reduced, but thermal displacement correction accuracy deteriorates
Solution Approach 1:
The system optimizes the parameter of sensor placement by using thermal displacement amount calculations to identify machine elements with significant thermal influence. The machine learning model analyzes thermal patterns and determines that sensors should be placed on specific high-influence elements, transforming the sensor installation strategy from arbitrary or comprehensive coverage to targeted, optimized placement.
3Measurement precision
If structural analysis by finite element method is performed to obtain optimum sensor installation positions, then thermal displacement correction accuracy is improved, but device complexity and difficulty of operation increase
Solution Approach 1:
The patent replaces complex mechanical/structural analysis methods (finite element method) with a machine learning-based computational approach. Instead of performing detailed structural analysis to determine thermal behavior, the system uses machine learning models that process thermal displacement data to identify optimal sensor positions, substituting complex physics-based analysis with data-driven pattern recognition.
Solution Approach 2:
The control device performs the optimization analysis automatically using the machine tool's own operational data. The system collects thermal displacement measurements during normal operation and uses this data to self-determine optimal sensor installation positions, eliminating the need for external expert analysis or complex manual calculations.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables simple and effective thermal displacement correction of upper and lower guides, improving machining precision while optimizing sensor placement to avoid wastefulness and complexity.
Implementation Method 1
the thermal expansion coefficients of these machine elements differ from each other. Therefore, due to factors such as a change in the temperature of the environment, there is concern over a plurality of machine elements thermally deforming
Data Source
Figure 1
Figure 2
Figure 3
AI summary
To provide a wire electric discharge machine capable of suitably and simply performing thermal displacement correction on upper and lower guides. Provided are a storage unit (21) that stores temperatures of machine elements and actual values for relative positions of upper and lower guides to be associated with each other as associated data; and a relational expression calculation unit (22) that infers and calculates a relational expression between the temperature of the machine element and the relative positions of the upper and lower guides by way of machine learning with this associated data as training data. Additionally provided are a position estimation unit (23) that substitutes temperatures of the machine element into the relational expression and calculates an estimated value for the relative position of the upper and lower guides; and a correction amount calculation unit (24) that calculates a correction amount for the upper and lower guides, based on the estimated value for the relative position. Further provided is a correction execution unit (25) that performs correction of the relative position of the upper and lower guides based on this correction amount.