Switchgear Infrared Hotspot Detection Using Synthetic AI Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for monitoring switchgear health, particularly for detecting hot spots, are inadequate as they require precise calibration and human intervention, and lack a standardized solution for different types and geometries of circuit breakers, with no affordable system available for early warning and monitoring.
Innovation Solution
An apparatus and system utilizing a machine learning classifier algorithm, trained on synthetic images generated by a Generative Adversarial Network, to analyze infrared images of switchgear for anomalous hot spots, eliminating the need for human calibration and region definition, and accommodating various circuit breaker geometries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional calibration methods are used to detect hot spots in switchgear, then measurement precision can be achieved for specific equipment, but the device complexity and need for human intervention increase, and the solution cannot be generalized to different types and geometries of circuit breakers
Solution Approach 1:
The patent uses synthetic images generated by GANs to create virtual copies of infrared images of switchgear. These synthetic copies allow the machine learning model to be trained on diverse equipment geometries and types without requiring physical calibration for each specific piece of equipment, thus reducing device complexity while maintaining detection accuracy
Solution Approach 2:
The patent transforms the detection approach by changing from calibration-based parameter adjustment to machine learning-based pattern recognition. The system learns temperature distribution patterns and hot spot characteristics directly from training data, eliminating the need for manual calibration parameters and making the solution adaptable to different equipment types
2Measurement precision
If manual calibration and region definition are required for hot spot detection, then measurement precision is maintained, but productivity and automation level decrease due to human intervention
Solution Approach 1:
The machine learning model performs self-service by automatically learning to identify hot spots and their characteristics from training data. The system autonomously detects anomalies without requiring human operators to manually calibrate equipment or define regions of interest, thereby improving productivity while maintaining precision
Solution Approach 2:
The patent applies preliminary action by pre-training the machine learning model on extensive synthetic data before deployment. This preliminary training enables the system to automatically recognize hot spot patterns in real operations without requiring manual intervention during actual monitoring, thus enhancing both precision and productivity
3Ease of operation
If a standardized solution is attempted for all switchgear types, then ease of operation improves, but measurement precision deteriorates due to geometric variations
Solution Approach 1:
The patent achieves universality by training the machine learning model on synthetic images representing multiple switchgear types and geometries. The single trained model can detect hot spots across different equipment configurations without requiring type-specific calibration, thus providing both ease of operation and maintained precision through its ability to handle geometric variations
Solution Approach 2:
The system employs dynamics by adapting its detection parameters automatically based on the input image characteristics. The machine learning model dynamically adjusts to different switchgear geometries and types encountered in real operations, maintaining measurement precision while providing a standardized ease of operation across diverse equipment
4Adaptability or versatility
If extensive training data from real switchgear is collected for machine learning, then adaptability improves, but loss of time and resources increase due to data collection requirements
Solution Approach 1:
The patent uses synthetic image generation by GANs to create copies of real infrared images with varied characteristics. This approach provides extensive diverse training data without requiring physical collection from multiple switchgear units, thus improving adaptability while eliminating the time loss associated with real data collection
Solution Approach 2:
The system performs preliminary action by pre-generating large datasets of synthetic training images before model training. This preliminary data preparation enables comprehensive adaptability training without the need for time-consuming real-world data collection during deployment, accelerating the overall implementation process
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
Enables accurate and automated detection of hot spots in switchgear and electrical components across different situations and types, reducing costs and improving monitoring efficiency without human intervention.
Implementation Method 1
an infra-red camera is configured to acquire the monitor infra-red image of the switchgear
Data Source
Figure 1~2
AI summary
The present invention relates to an apparatus for monitoring a switchgear. The apparatus comprises an input unit, a processing unit, and an output unit. The input unit is configured to provide the processing unit with a monitor infra-red image of a switchgear. The processing unit is configured to implement a machine learning classifier algorithm to analyse the monitor 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 on the basis of a plurality of training images. The plurality of training images comprises at least one synthetic image generated by an image processing algorithm. The output unit is configured to output information relating to the one or more anomalous hot spots.