Fault Isolation Reasoning for Diagnostic Ambiguity Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fault detection systems in complex engineered systems face challenges in isolating the cause of faults and predicting failures due to ambiguity in diagnostic indications, especially when sensor measurements are limited, leading to difficulties in distinguishing between multiple fault conditions.
Innovation Solution
A fault isolation and ambiguity resolution system utilizing analytic engines and a reasoning system to detect faults, identify ambiguity groups, and resolve ambiguities through evidence-based diagnosis, including the use of a single fault isolator, inference system, and fuzzy belief mapping to determine confidence levels and isolate faults within ambiguity groups.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple sensors are used to monitor system conditions, then fault detection capability is improved, but diagnostic ambiguity increases due to limited sensor measurements
Solution Approach 1:
The patent introduces a reasoning system with an inference engine as an intermediary between sensor data and fault diagnosis. This intermediary processes sensor measurements, applies domain knowledge, and resolves ambiguities by inferring the most likely fault conditions even when sensor data is limited or ambiguous.
Solution Approach 2:
The patent replaces traditional mechanical/direct observation-based diagnosis with an intelligent information processing system. The reasoning system uses software-based inference engines and knowledge bases to substitute for physical inspection and interpretation, enabling automated fault isolation and ambiguity resolution.
2Reliability
If traditional fault detection methods are used, then general fault conditions can be declared, but isolation of specific fault causes becomes challenging
Solution Approach 1:
The patent segments the fault diagnosis process into distinct functional components: sensor data acquisition, reasoning system processing, ambiguity group identification, and fault isolation. This segmentation allows each component to specialize in specific tasks, improving overall fault isolation capability while maintaining systematic monitoring.
Solution Approach 2:
The reasoning system acts as an intermediary that bridges general fault detection and specific fault isolation. It takes sensor inputs, applies inference rules, and produces isolated fault diagnoses, effectively mediating between the detection stage and the isolation stage.
3Measurement precision
If a reasoning system with ambiguity resolution is implemented, then diagnostic accuracy is improved, but system complexity increases
Solution Approach 1:
The reasoning system is designed as a universal platform that handles multiple fault types, sensor configurations, and diagnostic scenarios through a single integrated architecture. The inference engine and knowledge base can be applied across different system contexts, reducing the need for multiple specialized systems.
Solution Approach 2:
The system incorporates self-service capabilities where the reasoning system automatically resolves ambiguities and isolates faults without requiring external expert intervention. The inference engine autonomously processes data, applies rules, and generates diagnoses, reducing operational complexity despite increased structural complexity.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
A fault isolation and ambiguity resolution system (122) includes one or more analytic engines (206) executable by a processing system (130) and a reasoning system (500). The one or more analytic engines (206) are operable to detect a fault associated with a monitored system (100) based on data extracted from one or more data repository (120). The reasoning system (500) includes a single fault isolator (504) operable to identify an ambiguity group including the fault and one or more related faults of the monitored system. The reasoning system (500) also includes an inference system (512) operable to utilize evidence to resolve ambiguity between the fault and the one or more related faults of the ambiguity group as a diagnosis result.