Configurable Fault Trees for Transparent Electrical Fault Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fault analysis and prediction systems for electrical equipment, such as high voltage transformers, are proprietary and operate as 'black boxes', preventing customers from customizing and updating the underlying engineering models, which are based on years of expert knowledge.
Innovation Solution
A configurable fault tree structure that allows adjusting probability values based on detected states and observed faults, enabling customers to customize and update the model dynamically, using a probabilistic approach that incorporates Bayesian networks and objective data for failure mode analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If proprietary formulas and expert rules are used for fault analysis, then the reliability of fault prediction is improved, but the adaptability and transparency of the solution deteriorate
Solution Approach 1:
The system implements feedback loops where fault tree analysis results are continuously refined based on actual equipment performance data and observed faults. Probability values are dynamically updated as new data becomes available, allowing the proprietary expert knowledge to evolve and adapt to specific customer environments while maintaining the core reliability benefits of the original models.
Solution Approach 2:
The fault tree structure transitions from a static proprietary model to a dynamic configurable system where probability values and tree structures can be adjusted based on real-time data. This allows the system to maintain the reliability of expert-derived base models while adapting to specific customer needs and observed equipment behavior through continuous reconfiguration.
2Reliability
If proprietary models are deployed as black box products, then the reliability of fault analysis is improved, but the loss of information and transparency deteriorate
Solution Approach 1:
The system introduces an intermediary layer between the proprietary expert models and the customer data. This layer consists of configurable fault tree structures and probability adjustment mechanisms that translate opaque proprietary formulas into transparent, interpretable analysis results. Customers can observe how their specific equipment data flows through the fault tree and influences predictions, maintaining model reliability while eliminating the black box effect.
3Adaptability or versatility
If configurable fault tree structures are implemented, then the adaptability and customization capability are improved, but the device complexity increases
Solution Approach 1:
The fault tree system is segmented into modular, independently configurable components. Each fault tree represents a specific failure mode and can be configured separately with its own probability values and structure. This segmentation allows customers to customize only the relevant portions of the system without managing the complexity of the entire fault analysis framework, reducing the perceived complexity while maintaining adaptability.
4Productivity
If dynamic updating of probability values is enabled, then the productivity and responsiveness of fault assessment are improved, but the device complexity increases
Solution Approach 1:
The system implements self-service capabilities where the fault tree automatically updates probability values based on incoming equipment data without requiring manual reconfiguration. The configuration stored in memory enables the system to autonomously adjust its analysis parameters, improving productivity by eliminating manual intervention while containing complexity within the automated update mechanism rather than exposing it to users.
Data Source
AI summary
Embodiments are disclosed for adjusting at least one probability value for a subset of interconnected nodes of a configurable fault tree structure. The at least one probability value is associated with a state of at least one component of a first electrical equipment, and the configurable fault tree structure represents at least one failure mode relating to the at least one component. The adjusting is based on at least one of a detected state of the at least one component of the first electrical equipment and observed faults in a plurality of electrical equipment. Thereafter, generating an indication of the state of the at least one component based on the one or more adjusted probability values.