Machine Tool Dimension Prediction Using Driving State Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine tool technologies, such as those using neural networks and fuzzy inference, struggle to stably predict machining dimensions due to varying machining conditions and tool states, leading to inconsistent workpiece quality.
Innovation Solution
A machine tool machining dimensions prediction device comprising a driving state information acquirer, feature amount extractor, data analyzer, and machining quality prediction model generator, which collects and analyzes data from sensors to generate a regression-based prediction model for stable machining dimension prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If neural network learning is used to predict machining dimensions, then prediction capability is improved, but prediction stability deteriorates when number of tool types increases
Solution Approach 1:
The patent segments the machining dimension prediction into two distinct phases: a learning phase where the neural network is trained offline to establish the relationship between machining conditions and dimensions, and a prediction phase where the trained model is applied to actual machining. This segmentation allows the complex learning process to be completed once, ensuring stable and reliable predictions during subsequent machining operations without requiring continuous relearning.
2Adaptability or versatility
If neural network parameters are constantly updated to adapt to varying conditions, then adaptability is improved, but machining condition stability deteriorates
Solution Approach 1:
The patent performs the adaptation action in advance during the learning phase, where the neural network parameters are trained and optimized using historical machining data. Once the learning is complete, the parameters are fixed and not updated during actual machining operations. This preliminary action ensures the model adapts to varying conditions without compromising the stability of machining conditions during production.
3Adaptability or versatility
If fuzzy inference is used to correct machining conditions, then condition adjustment capability is improved, but prediction reliability deteriorates
Solution Approach 1:
The patent extracts and separates the prediction function from the condition correction function. The neural network is dedicated solely to predicting machining dimensions based on input conditions, while condition corrections are handled separately. This extraction ensures that the prediction model's reliability is not compromised by the uncertainties inherent in fuzzy inference-based condition adjustments.
Data Source
AI summary
A machine tool machining dimensions prediction device (100) includes: a data collector (10) to acquire driving state information of a machine tool; a feature amount extractor (211) to extract a feature amount from the driving state information; a data analyzer (311) to analyze the extracted feature amount; and a machining quality prediction model generator (312) to generate, from the analyzed information, a prediction model of a machining dimension of a workpiece. The machine tool machining dimensions prediction device (100) applies the feature amount and the driving state information to the prediction model during machining of the workpiece to predict a machining quality and refers to a machining dimension quality regulation to determine whether the machining quality satisfies a standard.