OPTIMIZATION OF FAULT DIAGNOSIS IN HELICOPTER HEALTH AND USA(16)

曝光台 注意防骗 网曝天猫店富美金盛家居专营店坑蒙拐骗欺诈消费者

2.5. OBJECTIVES
Information Technology (IT) solutions providing the infrastructure for the health monitoring functions will be reviewed, and improvement recommen-dations will be made to avoid IT related problems becoming a limited factor for the performance of the system.
Improved prognosis based on more precise load cycle calculation is cur-rently an important area of research. This path will however not be perused by this study. Nor will it treat problems related to airborne hardware, such as sensor and harness susceptibility to damage, digital hardware obsoles-cence, or practical problems related to the implementation of established usage monitoring techniques.
All tools used in this study, such as wavelets, arti.cial neural networks, and programming models are used without introduction. For any details on these technologies, the user might refer to the appendices and references.
Chapter 3

 

Current and Emerging Technologies
3.1 Introduction
This chapter explains the technologies that make up a HUMS. The state of the art for these technologies is reviewed, including an review of existing commercial solutions. From this, shortfalls for complying with the objectives of this study are identi.ed. Finally, improvement potential for the existing solutions are derived, and a number of research areas recommended.
The HUMS diagnosis logic accepts a set of sensor signals and produces a diagnosis of the underlying assets based on this information. This requires a set of formal steps, including contextual validation and correction, feature extraction, and classi.cation (Fig. 3.1).
Contextual
validation
and
correc-
tion is necessary in order to ensure that the data is representative for the state of the underlying assets. Any invalid data, like overly noisy data or data recorded in unfavorable conditions are removed or corrected at this stage. Such correction can be performed both before and after the feature extraction.
Feature extraction is to extract metrics about the system input which is more informative the evaluating at the raw input itself. The purpose of this step is to extract the essential characteristics of this input, so that it is more easily interpretable for the classi.er. The classi.er, for instance a fuzzy logic system or a neural network, is responsible for translating a set of features to an output diagnosis. As a classi.er is no more than a mapping tool, its performance is no more consistent than the features presented to it. It is thus vital that the pre-processing steps, contextual correction and feature extraction, does a good job in extraction features which makes it easy to
33

distinguish the di.erent classes, i.e. states of the underlying assets, that the classi.er is supposed to recognize.
A classi.er can be implemented as a neural network, fuzzy logic system, or simply a threshold tester. The classi.er accepts the data generated by the feature extractor, and makes a decision on the state of the monitored asset based on this. As a minimum, the classi.er must be able to distinguish assets in a normal condition from those behaving abnormally. In a more complex setting, the classi.er can produce more detailed information such as fault recognition and expected time to failure.
　
中国航空网 www.aero.cn
航空翻译 www.aviation.cn
本文链接地址：OPTIMIZATION OF FAULT DIAGNOSIS IN HELICOPTER HEALTH AND USA(16)