Abnormality Diagnosis Using Simulated Multi-Detector Time-Series
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional abnormality diagnostic methods face challenges in diagnosing issues when there is limited information available, especially when the number of sensors for data collection is large, or when past malfunction data is absent.
Innovation Solution
An abnormality diagnostic device that uses simulation results and time-series observation data from multiple detectors to diagnose abnormalities without relying on past malfunction information, employing techniques like complex power cepstrum and Dynamic Time Warping to identify patterns and classify abnormalities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional abnormality diagnostic methods are used that rely on accumulated malfunction information, then diagnostic accuracy can be maintained when sufficient data is available, but diagnostic capability deteriorates when information is limited or past malfunction data is absent
Solution Approach 1:
The patent performs preliminary simulation of abnormality occurrence before actual diagnosis. By pre-calculating expected observation results for various abnormality types through simulation, the system prepares diagnostic templates in advance that can be directly compared with real-time sensor data, eliminating the need to rely on accumulated historical malfunction information
Solution Approach 2:
The patent creates virtual copies of abnormality scenarios through simulation. Instead of relying on real historical malfunction data, the system generates simulated observation results that replicate what sensors would detect under various abnormal conditions, providing a comprehensive diagnostic reference library without requiring actual past failure data
2Measurement precision
If the number of detectors for acquiring information is increased, then measurement precision and diagnostic coverage are improved, but the amount of information that must be accumulated and processed increases significantly
Solution Approach 1:
The system pre-processes and simulates expected observation results for multiple detectors under various abnormality conditions before actual diagnosis. By calculating what each detector should observe during different failure modes in advance, the system creates a structured reference framework that reduces the complexity of processing real-time multi-detector data
Solution Approach 2:
The patent divides the diagnostic process into separate simulation tasks for each abnormality type and detector combination. By segmenting the overall diagnostic problem into individual simulation cases, the system can process and compare results in a modular fashion, making manageable even when numerous detectors are involved
3Reliability
If simulation of multiple abnormality types is performed for each detector, then diagnostic accuracy is improved by covering all possible failure modes, but computational complexity and processing time increase
Solution Approach 1:
The patent performs all simulation calculations for multiple abnormality types and detector combinations in advance, before actual diagnosis is needed. By pre-computing the entire library of expected observation results for various failure modes, the system transforms a computationally intensive real-time problem into a simple pattern-matching comparison between current sensor data and pre-simulated templates
Solution Approach 2:
The system performs comprehensive simulation for all possible abnormality types and detector combinations, even exceeding what might be strictly necessary. This excessive action ensures complete coverage of all failure modes and creates a robust reference library that simplifies real-time diagnosis despite the increased upfront computational effort
Data Source
AI summary
An abnormality diagnostic device includes a diagnoser configured to diagnose a type of abnormality that occurs in an abnormality diagnostic target on the basis of differences between abnormality simulation results for each type of abnormalities obtained by simulating a plurality of types of abnormalities in the abnormality diagnostic target and a plurality of time-series observation results obtained by observing the abnormality diagnostic target in time series using a plurality of detectors.