Optimized Fuzzy C-Means Clustering and Functional Principal Components for Post-Processing Dynamic Scenarios in the Reliability Analysis of a Nuclear System
Keywords
Statistical learning
Code:
24/2009
Title:
Optimized Fuzzy C-Means Clustering and Functional Principal Components for Post-Processing Dynamic Scenarios in the Reliability Analysis of a Nuclear System
Date:
Tuesday 25th August 2009
Author(s):
Di Maio, Francesco; Secchi, Piercesare; Vantini, Simone; Zio, Enrico
Download link:
Abstract:
This paper deals with the processing of accident scenarios generated from a dynamic reliability analysis of a Nuclear Power Plant (NPP). A large number of scenarios are simulated to account for the influence of the timing and magnitudes of fault events on the accident end states; post-simulation processing is then required for retrieving the safety-relevant information.
For classifying the final system state reached at the end of the accident scenarios, Fuzzy C-Means clustering is performed with different sets of Functional Principal Components (FPCs) of a selected relevant process variable. The approach allows capturing the characteristics of the process evolution determined by the occurrence,
timing, and magnitudes of the fault events.
An illustrative case study is considered, regarding the fault scenarios of the digital I&C
of the Lead Bismuth Eutectic eXperimental Accelerator Driven System (LBE-XADS).
The results obtained are compared with those of the Kth Nearest Neighbor (KNN) and Classification and Regression Tree (CART) classifiers.