Asset Degradation Path Classification for RUL Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current prognostic methods fail to accurately estimate the remaining useful life (RUL) of assets due to the assumption of a clear failure threshold, which is often vague in real-world applications with complex failure modes, and lack flexibility in handling multiple degradation signals and expert opinions.
Innovation Solution
The Path Classification and Estimation (PACE) method and system classify asset degradation based on exemplar data and estimate RUL by combining class memberships, allowing for flexible handling of various degradation signals and thresholds, and enabling accurate RUL estimation without relying on a single failure threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a clear failure threshold is assumed for degradation signals, then RUL estimation can be performed using simple threshold-based methods, but the estimation accuracy deteriorates when failure modes are complex and thresholds are vague
Solution Approach 1:
The patent transforms the degradation problem from threshold-based classification to continuous path-based classification by changing the parameter from discrete failure threshold to continuous degradation trajectory, enabling accurate RUL estimation for complex failure modes without requiring clear threshold definitions
Solution Approach 2:
The patent creates exemplar degradation paths from historical data that serve as reference templates, copying successful degradation patterns to classify and predict the RUL of current assets, thereby avoiding the need for explicit failure threshold definitions
2Adaptability or versatility
If multiple degradation signals and expert opinions are integrated into the prognostic model, then the flexibility and comprehensiveness of the model improve, but the model complexity increases
Solution Approach 1:
The patent segments the complex prognostic problem into distinct degradation paths that can be independently classified and analyzed, allowing multiple degradation signals to be processed through separate path templates without requiring a monolithic complex model
Solution Approach 2:
The patent creates a universal path classification framework that can handle multiple types of degradation signals and incorporate expert opinions through a unified exemplar-based approach, making the system versatile without proportionally increasing complexity
Data Source
AI summary
Path classification and estimation method and system used in combination with a computer and memory for prognosticating the remaining useful life of an asset by classifying a current degradation path of a current asset as belonging to one or more of previously collected degradation paths of exemplary assets and using the resulting classifications to estimate the remaining useful life of the current asset.