Industrial Mechanism Abnormality Estimation from Unit Phenomena
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately estimate abnormal phenomena and their causes in industrial mechanisms without relying on human experience or intuition, as feature identification is difficult due to varying conditions such as mechanism, command, and setting conditions.
Innovation Solution
An abnormality information estimation system that identifies multiple unit phenomena from operation data using analytical and machine learning methods, independent of mechanism type, to estimate abnormal phenomena and their causes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated abnormality estimation is implemented without relying on human experience, then productivity and consistency are improved, but measurement precision and reliability deteriorate due to difficulty in identifying features under varying conditions
Solution Approach 1:
The patent segments the abnormal phenomenon into multiple unit phenomena (e.g., vibration, temperature, sound) that can be independently identified and analyzed. This segmentation allows the system to handle complex abnormalities by breaking them down into manageable components, improving both automation capability and estimation accuracy.
Solution Approach 2:
The patent changes the parameters used for abnormality detection by identifying unit phenomena based on multiple operation data sets under different conditions (different mechanisms, commands, and settings). This approach enables the system to maintain high measurement precision across varying operational contexts.
2Adaptability or versatility
If multiple operation data sets under different conditions are used for training, then adaptability is improved, but device complexity increases due to multiple analysis models required
Solution Approach 1:
The patent creates a universal analysis model that can handle multiple types of mechanisms and operational conditions through a common framework. The model identifies unit phenomena that are applicable across different mechanisms, commands, and settings, eliminating the need for separate specialized models for each condition while maintaining high adaptability.
3Ease of operation
If unit phenomena are identified independently of mechanism type, then ease of operation is improved, but loss of information occurs due to ignoring mechanism-specific characteristics
Solution Approach 1:
The patent applies local quality by identifying unit phenomena at a general level that is independent of mechanism type, while still capturing mechanism-specific characteristics through the selection and combination of relevant unit phenomena. Each mechanism type can be analyzed using the same framework but with appropriate local adjustments in which unit phenomena are considered.
Data Source
AI summary
An abnormality information estimation system includes processing circuitry that identifies, based on operation data related to an operation of an industrial device that controls a mechanism, multiple unit phenomena due to the operation, and estimates abnormality information about an abnormality occurring in the mechanism based on the multiple of unit phenomena.