Machine Learning Feed Control for Crosscut Grooving Vibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine tools experience vibrations during crosscut grooving, leading to reduced machining accuracy and tool longevity, and current countermeasures are inefficient and operator-dependent, requiring extensive trial and error to find optimal feeding speeds.
Innovation Solution
A controller with a machine learning device that observes and learns the correlation between feeding amounts and vibration data, using state variables and determination data to determine optimal feeding speeds, reducing vibrations and improving machining precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a small feeding amount (low feeding speed) per unit cycle is specified to prevent rapid change in deflection, then vibrations are reduced, but cycle time increases
Solution Approach 1:
The system dynamically changes the feeding amount parameter based on real-time vibration measurements and machine learning predictions. The feeding amount is adjusted within a range rather than fixed, allowing optimization between vibration control and productivity. The control unit varies the feeding amount according to learned patterns of vibration behavior under different cutting conditions.
Solution Approach 2:
The system implements closed-loop feedback by measuring actual vibration amounts during cutting, comparing them against predicted values, and using this information to adjust the feeding amount. The vibration measurement unit continuously monitors the cutting part, and this feedback is fed into the machine learning device to refine predictions and guide feeding amount adjustments.
2Manufacturing precision
If feeding amount is reduced to prevent vibrations, then machining quality improves, but operator experience and trial-and-error time increase
Solution Approach 1:
The system performs self-optimization by automatically determining the appropriate feeding amount based on real-time vibration data and machine learning algorithms. The machine learning device learns from accumulated data and autonomously adjusts parameters without requiring operator intervention or expertise. The control unit executes these automated decisions, making the system self-sufficient in optimizing cutting parameters.
Solution Approach 2:
The system replaces operator judgment and manual adjustment with an automated control system based on machine learning and real-time vibration sensing. Instead of relying on operator experience to manually tune feeding amounts, the electronic control system automatically determines optimal parameters based on measured vibration data and learned patterns.
3Productivity
If feeding amount is increased to reduce cycle time, then productivity improves, but vibrations increase and machining accuracy deteriorates
Solution Approach 1:
The system transitions from static, pre-programmed feeding amounts to dynamic, real-time adjustment of the feeding amount. The feeding amount is continuously adapted based on current vibration conditions and machine learning predictions, allowing the system to optimize productivity while maintaining accuracy. This dynamic approach enables higher feeding amounts when conditions permit and reduces them when vibrations occur.
4Ease of manufacture
If conventional fixed feeding amount is used, then programming is simple, but vibrations occur at workpiece exit and accuracy is reduced
Solution Approach 1:
The system performs preliminary learning and prediction before critical cutting phases. The machine learning device accumulates vibration data and learns patterns in advance, enabling it to predict vibration risks before the tool exits the workpiece. This allows proactive adjustment of the feeding amount to prevent vibrations rather than reacting to them after they occur.
Data Source
AI summary
A machine learning device of a controller observes, as state variables that express a current state of an environment, feeding amount data indicating a feeding amount per unit cycle of a tool and vibration amount data indicating a vibration amount of a cutting part of the tool when the cutting part of the tool passes through the workpiece. In addition, the machine learning device acquires determination data indicating a propriety determination result of the vibration amount of the cutting part of the tool when the cutting part of the tool passes through the workpiece. Then, the machine learning device learns the feeding amount per unit cycle of the tool when the cutting part of the tool passes through the workpiece in association with the vibration amount data, using the state variables and the determination data.