Fault Tree Priority Scoring by Use Environment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fault tree analysis methods, such as Bayesian networks, require significant man-hours to create models for various fields, products, and use environments, and do not consider the influence degree of events on the entire product, making it inefficient for determining failure factor priorities.
Innovation Solution
A device and method for calculating failure factor priorities using a fault tree that considers the field, product, use environment, and influence degree by analyzing past defect information to assign scores to events based on co-occurrences and relationships, without constructing a comprehensive Bayesian network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Bayesian network is used to express defect causal relationship and determine event priorities, then the priority determination can be performed, but significant man-hours are required to create models for various fields, products, and use environments
Solution Approach 1:
The patent segments the fault tree analysis by separating the event priority calculation from the comprehensive Bayesian network model. Instead of creating a complete Bayesian network for the entire system, the patent calculates priorities for individual events and intermediate events independently using simplified probability calculations based on end event priorities and logical relationships (AND/OR gates), thereby reducing model creation time while maintaining priority determination accuracy
Solution Approach 2:
The patent extracts the essential function of priority determination from the complex Bayesian network framework. By taking out only the necessary probability calculation components and applying them selectively to end events and intermediate events in the fault tree, the patent achieves priority determination without requiring the full complexity of a Bayesian network model, thus reducing man-hours while preserving measurement precision
2Adaptability or versatility
If comprehensive Bayesian network model is created for all fields, products, and use environments, then complete coverage is achieved, but the complexity and man-hours required increase significantly
Solution Approach 1:
The patent creates a universal fault tree analysis method that can be applied across different fields, products, and use environments without requiring separate Bayesian network models for each. The same simplified probability calculation approach works universally for determining event priorities in any fault tree structure, providing adaptability and versatility while avoiding the complexity of creating customized comprehensive models for each application scenario
Solution Approach 2:
The patent applies local quality by focusing computational resources on calculating priorities for specific end events and intermediate events that are relevant to the analysis at hand, rather than creating a comprehensive model for all possible events. This allows the method to adapt to different fields and products by calculating only the locally necessary priorities, reducing overall model complexity while maintaining coverage where needed
3Productivity
If traditional fault tree analysis is performed without considering use environment, then the analysis can be completed quickly, but the priority evaluation does not reflect real-world failure likelihood
Solution Approach 1:
The patent incorporates use environment information and end event priority data in advance into the calculation framework before performing the fault tree analysis. By preliminarily gathering information about operational conditions, failure rates in different environments, and criticality of end events, the patent can quickly calculate accurate priorities for intermediate events without sacrificing analysis speed, as the environmental factors are already integrated into the probability calculations
Data Source
AI summary
Automatically generating a fault tree on the basis of the causal relationship of defects to events to be analyzed, and calculating a priority degree (score) for the individual events in the generated fault tree by means of the number of co-occurrences of “events” and “event-related information” in past defect information on the basis of “event-related information” in which a component is used. A scored fault tree to which the scores have been applied is presented.