Machine Tool Thermal Displacement Control With Learned Correction Coefficients
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine tools face challenges in accurately correcting thermal displacement errors caused by thermal expansion, which affect the accuracy of machining due to the complexity of heat sources and the difficulty in accurately calculating thermal displacement corrections.
Innovation Solution
A controller and machine tool system that uses a plurality of temperature sensors to calculate a correction coefficient for a thermal displacement correction equation, incorporating a learning function to adjust the correction coefficients based on machining data and dimension measurement data, ensuring precise thermal displacement correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple temperature sensors are used to measure thermal displacement correction amounts, then the accuracy of thermal displacement correction is improved, but the device complexity increases
Solution Approach 1:
The system uses actual machining results (dimension measurement data) as feedback to learn and optimize the correction coefficients. The learning unit continuously adjusts the correction coefficients based on the relationship between temperature sensor readings and actual machining dimensions, improving accuracy without requiring additional sensors.
Solution Approach 2:
The system performs self-optimization through machine learning, automatically adjusting its own correction coefficients based on accumulated machining data. This eliminates the need for manual calibration and reduces the burden on operators while maintaining high accuracy.
2Manufacturing precision
If correction coefficients are manually set based on temperature sensor data, then the manufacturing precision is improved, but the ease of operation deteriorates
Solution Approach 1:
The learning unit automatically optimizes correction coefficients using machine learning algorithms based on accumulated machining data. The system performs self-calibration without operator intervention, eliminating the need for manual coefficient setting while maintaining high machining accuracy.
Solution Approach 2:
The system replaces manual calibration operations with automated machine learning processes. Instead of operators manually adjusting coefficients based on experience and temperature data, the system uses computational algorithms to automatically determine optimal correction values.
3Manufacturing precision
If thermal displacement correction is applied using traditional methods, then the manufacturing precision is improved, but the loss of time increases due to manual calibration
Solution Approach 1:
The system performs preliminary learning during idle periods or between machining operations, accumulating data and optimizing correction coefficients in advance. This allows the system to be ready with optimized coefficients when needed, eliminating calibration time during production.
Solution Approach 2:
The learning process operates continuously in the background, constantly improving correction coefficients based on accumulated data. This continuous learning ensures that the system maintains optimal accuracy without requiring periodic shutdowns for manual recalibration.
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
The system effectively calculates and applies correction coefficients to maintain machining accuracy within tolerance ranges, reducing the occurrence of thermal displacement errors and ensuring precise machining dimensions.
Implementation Method 1
Components of a machine tool undergo thermal expansion due to heat generated by operating the machine tool and external heat around the machine tool.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention has an object to provide a device that automatically performs a thermal displacement correction in a machine tool. To realize this object, a controller according to the present invention includes: a machining data collector 67 configured to collect machining data including a date and a time when at least one machined portion that is a part of a workpiece and that was set as a learning target was machined by a machine tool; a temperature collector 29 configured to collect, at predetermined time intervals, temperature data obtained at a plurality of positions on the machine tool; a dimension data input receiver 75 configured to receive an input of dimension measurement data obtained by measuring a dimension of the at least one machined portion after the at least one machined portion has been machined; a learning data generator 85 configured to generate learning data including the machining data and the dimension measurement data; and a machine learning executor 95 configured to execute a machine learning based on the temperature data and the learning data, and configured to calculate a correction coefficient used in a thermal displacement correction equation for correcting a displacement caused by a change in a temperature of the machine tool.