Bearing Temperature Correlation for Early Abnormality Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting abnormalities in numerous target machines, such as bearings, are inefficient due to high false alarm rates caused by temperature and load changes, making it difficult to detect issues at an early stage, especially in remote locations.
Innovation Solution
An abnormality detection device that acquires drive and non-drive side temperatures and vibrations, uses correlation analysis to detect deviations, and determines operational states like stoppages, allowing for early anomaly detection without constant monitoring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a strict threshold value is set for vibration monitoring, then early abnormality detection is improved, but false alarms increase due to fluctuations from temperature and load changes
Solution Approach 1:
The patent changes the monitoring parameter from absolute vibration value to temperature correlation coefficient. By calculating the correlation between drive-side and non-drive-side bearing temperatures, the system detects abnormalities through deviations in temperature relationships rather than fixed vibration thresholds, thereby reducing false alarms while maintaining detection sensitivity
Solution Approach 2:
The patent introduces temperature correlation coefficient as an intermediary parameter. Instead of directly monitoring vibration values, the system uses temperature measurements from both sides of the bearing as intermediate indicators to infer abnormality, which reduces the impact of direct vibration fluctuations and environmental factors
2Measurement precision
If constant monitoring of trend data is implemented to detect early abnormalities, then detection accuracy is improved, but system complexity and resource requirements increase for managing numerous target machines
Solution Approach 1:
The patent applies partial monitoring by focusing only on the correlation relationship between two temperature parameters rather than continuously analyzing all vibration trend data. This selective approach maintains detection accuracy while significantly reducing computational complexity and resource requirements for managing multiple target machines
3Ease of operation
If threshold-based monitoring is used to manage numerous target machines, then system simplicity is maintained, but early abnormality detection capability is lost
Solution Approach 1:
The system automatically calculates temperature correlation coefficients and compares them against predetermined thresholds, enabling self-service operation without requiring constant manual intervention. This maintains ease of operation while improving detection capability through automated correlation analysis
Solution Approach 2:
The patent transforms the monitoring approach by changing from simple absolute value threshold comparison to correlation coefficient threshold comparison. This parameter transformation enables early abnormality detection while maintaining the simplicity of threshold-based operation through automated calculation
Data Source
AI summary
Provided is an abnormality detection device which detects an abnormality in a target machine, comprising: a first acquisition unit which acquires a drive side temperature of the target machine; a second acquisition unit which acquires a non-drive side temperature of the target machine; a correlation storage unit which stores a correlation between the drive side temperature and the non-drive side temperature based on the drive side temperature and the non-drive side temperature during normal operation of the target machine; a detection unit which detects a deviation from the correlation stored in the correlation storage unit on the basis of the drive side temperature acquired by the first acquisition unit and the non-drive side temperature acquired by the second acquisition unit; and, an output unit which outputs the deviation from the correlation which was detected by the detection unit as an abnormality in the target machine or as an abnormality indication.