Vibration and oil debris analysis for advanced helicopter health and usage monitoring systems

Requirements for future advances in helicopter Integrated Health and Usage Monitoring Systems (IHUMS) are discussed. Vibration and oil debris analysis for efficient IHUMS is presented. Mapping functions are deduced and implemented to transform primary features (such as vibration harmonics and element concentration levels in oil) to a set of features that yield better discrimination. Fault diagnosis is based on unsupervised and supervised machine learning methods. Neural network problems associated with identifying disjoint or complicated patterns are partially solved.