Switchgear Hot Spot Detection Using Synthetic IR Image Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for monitoring switchgear for hot spots are not cost-effective, require precise calibration, and rely heavily on manual analysis, which is time-consuming and expensive.
Innovation Solution
An apparatus comprising an input unit, a processing unit, and an output unit that utilizes a machine learning classifier algorithm trained on synthetic infra-red images generated from CAD drawings to detect anomalous hot spots in switchgear.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual analysis of infra-red images is used to detect hot spots, then detection capability is achieved, but time consumption and cost increase significantly
Solution Approach 1:
The patent replaces the manual mechanical analysis process with an automated image processing system that uses algorithms to detect hot spots in infra-red images. The system automatically identifies temperature anomalies and generates reports, eliminating the need for manual image analysis while maintaining detection accuracy.
Solution Approach 2:
The system enables self-service monitoring by automatically capturing infra-red images, processing them through analysis algorithms, and generating diagnostic reports without requiring manual intervention. The switchgear monitoring system performs self-diagnosis and alerts operators to potential issues autonomously.
2Measurement precision
If precise calibration is performed to measure temperature at the right position, then measurement accuracy is improved, but device complexity and calibration time increase
Solution Approach 1:
The patent creates a digital model or map of the switchgear's thermal characteristics that can be reused across multiple measurements. Instead of recalibrating the system for each measurement, the calibrated thermal model is copied and applied to subsequent infra-red image analyses, maintaining accuracy while eliminating repetitive calibration procedures.
Solution Approach 2:
The system performs calibration and creates thermal models in advance during the manufacturing or initial setup phase. This preliminary action ensures that the measurement system is pre-configured with the correct parameters and models, eliminating the need for complex on-site calibration while maintaining measurement precision.
3Device complexity
If a common solution is applied to all switchgear types, then device complexity is reduced, but adaptability to different switchgear geometries decreases
Solution Approach 1:
The patent implements a dynamic monitoring system that can adapt its parameters and analysis methods based on the specific type and geometry of the switchgear being monitored. The system automatically adjusts its thermal models, image processing parameters, and detection thresholds to match the characteristics of different switchgear configurations, maintaining both simplicity and adaptability.
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
The system enables accurate and automated detection of hot spots in switchgear, reducing the need for manual calibration and analysis, and allowing for real-time monitoring and early warning of potential failures.
Implementation Method 1
an infra-red camera is configured to acquire a monitored infra-red image of the switchgear
Data Source
AI summary
An apparatus for monitoring a switchgear includes: an input unit; a processing unit; and an output unit. The input unit is provides the processing unit with a monitored infra-red image of a switchgear. The processing unit implements a machine learning classifier algorithm to analyse the monitored infra-red image and determine if there is one or more anomalous hot spots in the switchgear. The machine learning classifier algorithm has been trained based on a plurality of different training images, the plurality of training images including a plurality of synthetic infra-red images generated by an image processing algorithm. The output unit outputs information relating to the one or more anomalous hot spots.