Machining Section Detection From Multi-Pattern Feed Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to accurately distinguish between non-machining and machining sections in tapping machining processes due to the presence of multiple signal patterns in the feed signal, leading to erroneous determination of machining sections.
Innovation Solution
A diagnostic system that includes a receiving unit to combine operation information from outward and return movements, an extraction unit to extract feature information from vibration data, and a selection unit to generate models using machine learning for accurate determination of machining sections by calculating likelihood based on reference feature information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the machining section is determined based on one pattern of the feed signal, then the determination process is simple, but the machining section is erroneously determined when multiple signal patterns are present
Solution Approach 1:
The determination apparatus segments the feed signal into multiple signal patterns (e.g., acceleration pattern, constant speed pattern, deceleration pattern) and creates separate determination units for each pattern. Each determination unit is specialized to handle its specific signal pattern, allowing accurate identification of machining sections despite the complexity of multiple patterns present in the feed signal.
2Measurement precision
If multiple determination units are used to handle different signal patterns, then the machining section is accurately determined, but the device complexity increases
Solution Approach 1:
The signal pattern identification unit serves as an intermediary that analyzes the feed signal and identifies which signal pattern is present. Based on this identification, it selects and activates only the appropriate determination unit, thereby avoiding the need to maintain all determination units simultaneously active and reducing overall system complexity.
Solution Approach 2:
The determination apparatus dynamically adapts its structure by activating only the determination unit corresponding to the currently detected signal pattern. This dynamic activation approach allows the system to maintain high determination accuracy while reducing the effective complexity at any given moment, as only one determination unit is actively processing at a time.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
A determination apparatus includes a receiving unit (201) to receive operation information corresponding to an action being performed by a machine (1) to be diagnosed and a detection signal of a physical quantity that changes according to the action of the machine (1); a taking-out unit (201) to take out, from the detection signal, an operation detection signal indicating that the machine (1) is operating, based on the operation information; an extraction unit (202) to extract feature information of the operation detection signal; a selection unit (203) to select, from the feature information, particular feature information to be compared with a plurality of reference feature information; and a determining unit (205) to determine a machining section of the machine (1) in the feature information, based on the plurality of reference feature information and the particular feature information.