Thermal Image Correlation Diagnosis for Device Abnormalities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing abnormality diagnosis technologies for devices, particularly focusing on temperature, lack accuracy in detecting anomalies.
Innovation Solution
A program and diagnostic device that utilize a computer to set regions in a temperature image, identify correlations between feature amounts in these regions, and perform abnormality diagnosis based on changes in these correlations, incorporating feature amounts related to temperature and operation state, and optionally using visible light images to enhance region setting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If temperature images are used for abnormality diagnosis, then non-contact measurement is achieved, but measurement precision is insufficient
Solution Approach 1:
The temperature image is divided into multiple regions of interest (ROIs) corresponding to different components or areas of the device. Each region's temperature characteristics are analyzed separately, and correlations between regions are computed to detect abnormalities. This segmentation allows precise local temperature analysis while maintaining non-contact measurement advantages.
Solution Approach 2:
The patent transforms the temperature image data into correlation parameters that represent thermal relationships between different regions. By computing correlation coefficients between temperature variations across regions, the system extracts meaningful diagnostic information that improves abnormality detection precision while maintaining non-contact measurement capability.
2Measurement precision
If multiple sensors are installed for accurate temperature measurement, then measurement precision improves, but device complexity increases
Solution Approach 1:
Instead of installing multiple physical temperature sensors across the device, the patent uses a single infrared camera to capture temperature images that copy the thermal state of the entire device. This optical copying approach provides comprehensive temperature data without the complexity of multiple contact sensors, maintaining measurement precision while reducing device complexity.
Solution Approach 2:
The patent replaces the mechanical sensor installation system with an optical measurement system. An infrared camera captures temperature distributions non-contactly, substituting the need for physical temperature sensors and their complex installation, wiring, and calibration with a simplified optical imaging approach that provides equivalent or superior measurement capability.
3Measurement precision
If correlation analysis between multiple regions is performed, then abnormality diagnosis accuracy improves, but computational complexity increases
Solution Approach 1:
The patent performs preliminary segmentation of the temperature image into meaningful regions based on device structure or thermal characteristics before correlation analysis. By pre-defining regions of interest and their expected thermal relationships, the system reduces the computational scope of correlation calculations while maintaining diagnostic accuracy, avoiding the need to compute correlations across all possible pixel combinations.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Accurately diagnoses device abnormalities by analyzing temperature distribution and correlations, reducing the need for extensive sensor installation and lowering costs while improving diagnostic precision and ease of use.
Implementation Method 1
an infrared camera that captures a temperature image showing a temperature distribution of a device
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are a program, a diagnostic method, and a diagnostic device, in which a computer performs first processing of setting a plurality of regions in a temperature image showing a temperature distribution of a device, second processing of identifying a change in a first correlation which is a correlation of a first feature amount relating to temperature between regions that are different from each other, and third processing of performing an abnormality diagnosis of the device based on the change in the first correlation.