Research Andapplication Of Signal Processing And Status Recognition For The Pipe

With the rapid development of aerospace technology, the reliability and safety ofaero-engine becomes the focus and attracts much attention. The pipeline system ofaero-engine, whose safe state plays an obvious role in the reliability and safety ofaero-engine. Thus, the study of aero-engine pipeline fault diagnosis is a key part ofensuring the normal operating of aero-engine. In the study of aero-engine pipelinefault diagnosis, the study of time-frequency algorithm is significant to the study offault feature, and the implementation of real time monitoring system can control thestate of pipeline rapidly. However, the study of engine pipeline fault monitoringsystem at home and abroad is still at the stage of development, hence, muchtheoretical analysis and system implementation is needed to support the study ofengine fault diagnosis.This paper takes the engine pipeline as the study object. In the theoreticalanalysis, the reliability of Fiber Brag Grating (FBG) while in signal acquisition isverified. Besides, three time-analysis algorithm—wavelet transform(WT)、EmpiricalMode Decomposition(EMD)、Intrinsic Time-scale Decomposition(ITD) are used inanalyzing of aero-engine pipeline signal and simulate signal. Decomposition EffectRatio(DER) is proposed for balancing the decomposition effect of those threealgorithm. Moreover, under the study of signal decomposition, the pipeline featureextraction is also researched, which provide a theoretical basis for Monitoring. In thedesign of monitoring system, using programming technology, the monitoring systemwith high speed data transforming, high efficiency analyzing and high real-timeperformance is implemented, and provides technical support for monitoring theworking condition of engine pipeline. The main of the paper is as follows:(1) According to the simulation of three time-analysis algorithm—wavelettransform(WT)、 Empirical Mode Decomposition(EMD)、 Intrinsic Time-scaleDecomposition(ITD) Decomposition Effect Ratio(DER) is proposed for balancing thedecomposition effect. The DER is proved to be efficacious. Moreover, time varianceis analyzed to present the application of the three algorithms in real-time aspects.(2) According to the conclusion in (1). we can get the change trend of distributedtemperature signal by extracting low frequency component with using the time-frequency algorithms. In view of the actual distributed strain signal,The DER isproved to be efficacious again by using the three algorithm analysis the fault signalrespectively,at the same time, relative energy ratio and kurtosis is extracted as afeature vector, and BP neural network recognition results verify the validity of themonitoring method, which provide a theoretical basis for monitoring system.(3) Client and service design mode is designed to realize concurrent processingof multiple clients online of aero-engine real-time monitoring system. Among themodules, the assembly line work style is adopted. By using the producer andconsumer module, the problem of data synchronism is solved. In data storing, thecache technology is adopted to improve the performance. The testing results of thesystem show that while in the state of multiple clients linking, high speed datatransferring, the signal can be processed real-time, original data can be stored,interactive interface can be responded quickly. Moreover, the offline signalprocessing ability can prevent the future fault.