Verification and Validation of System Health Management Models using Parametric Testing

Johann Schumann,Erik Reed,Ole Mengshoel

doi:10.2514/6.2011-1445

Abstract

System Health Management (SHM) systems have found their way into many safety-critical aerospace and industrial applications. A SHM system processes readings from sensors throughout the system and uses a Health Management (HM) model to detect and identify potential faults (diagnosis) and to predict possible failures in the near future (prognosis). It is essential that a SHM system, which monitors a safety-critical component, must be at least as reliable and safe as the component itself—false alarms or missed adverse events can potentially result in catastrophic failures. The SHM system including the HM model, a piece of software, must therefore undergo rigorous Verification and Validation (VV they need to be set carefully for reliable and accurate HM reasoning. We are investigating the use of Parametric Testing (PT), which uses a combination of n-factor and Monte Carlo methods, to exercise our HM model with variations of perturbed parameters. Multivariate clustering on the analysis is used to automatically find structure in the data set and to support visualization. Our approach can yield valuable insights regarding the sensitivity of parameters and helps to detect safety margins and boundaries. As a case study we use HM models from the NASA Advanced Diagnostics and Prognostics Testbed (ADAPT), which is a realistic hardware setup for a distributed power system as found in spacecraft or aircraft.

Highlights

System Health Management (SHM) systems are ubiquitious in many safety-critical aerospace and industrial applications, including major components of aircraft
While many different approaches and tools for FDDR and SHM exist,1–3 this paper focuses on techniques and tools using Bayesian networks
We investigate a real-world electrical power system known as the Advanced Diagnostics and Prognostics Testbed (ADAPT)

Summary

Introduction

System Health Management (SHM) systems are ubiquitious in many safety-critical aerospace and industrial applications, including major components of aircraft (e.g., jet engines, hydraulics, or electric power systems). During operation of the monitored system, the system processes readings from sensors throughout the system and use a Health Management (HM) model to detect and identify faults (diagnosis). Often such systems can trigger recovery activities. We will describe research which has been performed using Health Management models that are modeled using Bayesian networks (BNs), a very powerful modeling paradigm to express notions of cause and effect, probability, and reliability surpassing many traditional (discrete) approaches toward HM like table-driven techniques or fault trees. A low pressure reading in conjunction with vibration indicates a worn bearing (p=63.3%) Such networks, which can become very large, may be compiled into clique trees or arithmetic circuits for efficient embedded execution.

Background

Parametric Model Analysis

Clustering Analysis

Conclusions