Rotating-Member Abnormality Diagnosis Using Multi-Format Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing abnormality diagnosing systems for devices with rotating members lack accuracy in identifying multiple types of abnormality causes and have limited capability in distinguishing between various causes.
Innovation Solution
An abnormality cause identifying system that converts measurement data into multiple new-format data pieces, including frequency analysis, waterfall, Bode, polar, and shaft center trace data, and performs nondimensionalization using feature frequencies to analyze these data pieces in combination, enhancing accuracy and identifying a broader range of causes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If measurement data is converted to only one type of conversion data piece, then the system complexity is low, but the abnormality cause identification accuracy is insufficient and the number of identifiable abnormality causes is small
Solution Approach 1:
The patent applies segmentation by dividing the data conversion process into multiple independent conversion paths, where measurement data is converted into multiple distinct types of conversion data pieces (e.g., vibration data, temperature data, pressure data) simultaneously. Each conversion type analyzes different aspects of the device state, enabling comprehensive abnormality cause identification without requiring a single overly complex conversion system
Solution Approach 2:
The patent implements multi-functionality by designing a data processing system that performs multiple conversion functions in parallel. The same measurement data is processed to generate various types of conversion data pieces that can identify different abnormality causes, making the system capable of handling diverse diagnostic requirements through a unified multi-functional framework
2Adaptability or versatility
If multiple types of conversion data pieces are generated and analyzed in combination, then the number of identifiable abnormality causes increases, but the data processing complexity increases
Solution Approach 1:
The patent segments the analysis process by assigning each conversion data piece type to analyze specific abnormality cause categories. This segmentation allows the system to handle multiple abnormality types through structured, modular analysis paths, reducing the overall processing complexity while maintaining high versatility in identifying different causes
Solution Approach 2:
The patent merges multiple conversion data pieces through integrated analysis that combines information from vibration, temperature, pressure, and other data types. This merging approach enables the system to identify a broader range of abnormality causes by synthesizing complementary information from different data sources, achieving high versatility without proportionally increasing processing complexity
3Measurement precision
If conventional single-format data analysis is used, then the processing speed is fast, but the identification accuracy and range of abnormality causes are limited
Solution Approach 1:
The patent applies preliminary action by performing multiple data conversions and preparing various types of conversion data pieces in advance, before abnormality analysis is needed. This preliminary processing organizes measurement data into multiple ready-to-use formats, enabling fast retrieval and analysis when abnormalities occur, thus maintaining high processing speed while achieving accurate identification across multiple abnormality causes
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An abnormality cause identifying system 10 for a device including a rotating member includes: acceleration sensors 22a and 22b, a pickup sensor 24, and a temperature sensor 26, which observe a state of a rotating member R and acquire a measurement data piece; a measurement data piece converting portion 30 configured to convert the measurement data piece into two or more new-format conversion data pieces that are different from each other; and an abnormality cause identifying portion 40 configured to identify an abnormality cause of the device by analyzing the conversion data pieces created by the measurement data piece converting portion 30.