Classification analysis of vibration data from SH-60B Helicopter Transmission Test Facility

Abstract

The U.S. Navy is currently evaluating an integrated diagnostic system for its rotary wing aircraft. The system is referred to as the Health Usage and Monitoring Systems (HUMS). The program's objective is to develop an automated diagnostic system that can identify mechanical faults within the power train of helicopters using vibration analysis. This thesis uses data provided by the Helicopter Transmission Test Facility at the Naval Air Warfare center, Trenton, New Jersey. The goal of this thesis is to conduct data analysis to identify a fault within the helicopter test transmission using a tree-structured model. Prior to conducting tree analysis, an attempt is made to reduce the amount of data by principal component analysis. All statistical analysis was completed with S-Plus Software (MathSoft Inc., 1995)