Rotating Component Fault Detection via Reference Signal Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting faults in mechanical structures with rotating machine components can only distinguish between normal and developing fault conditions, failing to provide early warnings for specific fault conditions and differentiate between various fault types.
Innovation Solution
A method that utilizes an input signal representing mechanical vibrations, a reference signal generated based on parameters and rotational speed of the rotating machine component, and processing instructions to select among predetermined condition evaluations, enabling early detection and differentiation of fault conditions through signal processing and analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional vibration analysis methods are used to detect faults, then fault detection capability is provided, but the ability to differentiate between specific fault conditions and provide early warnings is insufficient
Solution Approach 1:
The method segments fault detection into multiple distinct fault condition categories (e.g., inner race fault, outer race fault, ball fault, gear mesh fault) by comparing vibration signals against multiple specialized reference signals, each tailored to detect specific fault types. This segmentation enables precise differentiation between various fault conditions rather than providing a single generic fault detection.
Solution Approach 2:
The method changes parameters by generating reference signals with specific frequency characteristics matched to different fault conditions. By adjusting the frequency and temporal parameters of reference signals based on rotational speed and fault type, the system can selectively detect and differentiate between various fault conditions in the vibration spectrum.
2Reliability
If generic fault detection methods are applied, then normal vs. developing fault distinction is achieved, but early warning capability for specific fault conditions is lost
Solution Approach 1:
The method performs preliminary action by continuously comparing vibration signals against multiple pre-configured reference signals that represent different fault conditions. This continuous comparison against specialized references enables early detection and warning of specific fault conditions before they progress to severe failures, providing timely alerts for maintenance planning.
Solution Approach 2:
The system implements feedback by continuously monitoring vibration signals, comparing them against reference signals for different fault conditions, and providing real-time condition evaluations. This feedback mechanism enables ongoing assessment of mechanical structure health and provides timely warnings when specific fault conditions are detected.
3Measurement precision
If multiple fault condition evaluations are implemented, then fault differentiation capability is improved, but system complexity increases
Solution Approach 1:
The method uses copying by creating reference signals that replicate the expected vibration patterns for different fault conditions. These reference signals serve as templates that are compared against actual vibration measurements, enabling fault differentiation without requiring complex analysis algorithms. The reference signals act as simplified copies of fault characteristics for direct comparison.
Solution Approach 2:
The system achieves universality by using a single processing framework that can evaluate multiple fault conditions simultaneously through comparison against different reference signals. This multi-functional approach allows the same basic comparison mechanism to detect various fault types (bearing faults, gear faults, etc.) without requiring separate dedicated systems for each fault type.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for providing a condition evaluation of a mechanical structure which includes a rotating machine component comprises providing an input signal that represents mechanical vibrations in the mechanical structure; and selecting, on the basis of the input signal and a number of reference signals, one among a plurality of predetermined condition evaluations for the mechanical structure. When the rotating machine component comprises a gear transmission, the condition evaluation is selected using amplitude demodulation. When the machine component comprises a rotating bearing, the condition evaluation is selected using cross-correlation.