Abnormality Diagnosis Using Multi-Length Mahalanobis Unit Spaces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional abnormality diagnosis methods using Mahalanobis distance struggle with accurately distinguishing between normal and abnormal variations in equipment data, particularly due to undefined data length of the unit space, leading to erroneous determinations of normal variations as abnormal and vice versa.
Innovation Solution
Creating multiple unit spaces with different data lengths based on correlation coefficients and normal variation cycles to calculate multiple Mahalanobis distances, allowing for more accurate abnormality determination by comparing these distances against thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single unit space with fixed data length is used for abnormality diagnosis, then the determination process is simple, but normal variations may be erroneously determined as abnormal or abnormal signs may be missed
Solution Approach 1:
The patent divides a single unit space into multiple unit spaces with different data lengths (e.g., 1-hour, 6-hour, 12-hour, 24-hour unit spaces). Each unit space segment captures variations at different time scales, allowing the system to distinguish between short-term normal fluctuations and long-term abnormal trends, thereby improving determination accuracy without requiring a single complex unit space structure
Solution Approach 2:
The patent adds the dimension of time scale by creating unit spaces at multiple temporal resolutions. By analyzing data at different time lengths (1h, 6h, 12h, 24h), the system transforms a one-dimensional analysis into a multi-dimensional approach, enabling better differentiation between normal and abnormal variations across different temporal perspectives
2Speed
If the data length of unit space is too short, then sudden changes can be detected quickly, but normal variations are erroneously determined as abnormal
Solution Approach 1:
The patent segments the data length into multiple levels (1-hour, 6-hour, 12-hour, 24-hour unit spaces). The 1-hour unit space provides quick response to sudden changes, while longer unit spaces filter out normal variations. By combining results from multiple segments, the system achieves both fast detection and high accuracy
Solution Approach 2:
The patent uses multiple unit spaces with different data lengths rather than a single optimal length. This partial approach (using several suboptimal unit spaces) compensates for the limitations of each individual unit space, where shorter ones detect sudden changes quickly but produce false positives, and longer ones filter false positives but slow down detection
3Reliability
If the unit space is updated frequently, then the diagnosis reflects current state, but computational load increases
Solution Approach 1:
The patent segments the update process by creating multiple unit spaces with different data lengths at different update intervals. Longer unit spaces are updated less frequently, reducing overall computational load while shorter unit spaces are updated more frequently to maintain timeliness. This segmented update strategy balances reliability and energy consumption
Data Source
AI summary
An abnormality diagnosis method for diagnosing an abnormality in operational state of a diagnosis subject includes creating a unit space from normal operation data of the diagnosis subject, the unit space serving as a reference for determining the operational state of the diagnosis subject, acquiring data having state quantities of a plurality of evaluation items from the diagnosis subject, calculating a Mahalanobis distance of the data acquired, using the unit space created, and determining an abnormality in the operational state of the diagnosis subject based on the Mahalanobis distance calculated. The creating a unit space includes creating a plurality of unit spaces having mutually different data lengths The calculating a Mahalanobis distance includes calculating a plurality of Mahalanobis distances using the plurality of unit spaces created. The determining an abnormality includes determining an abnormality based on the plurality of Mahalanobis distances calculated.