Font Size: a A A

Vibration Signal Analysis Based Fan Performance Evaluation And Testing

Posted on:2015-06-24Degree:MasterType:Thesis
Country:ChinaCandidate:Z Y QuFull Text:PDF
GTID:2272330452966826Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
Studying fan performance evaluation is of vital significance since itis widely used in many industry areas. Based on the analysis ofmechanical behavior and aerodynamic performance, vibration signalanalysis based fan performance evaluation method is developed.In order to study fan vibration characteristics in depth, flap-wisevibration model of fan blade is established and analyzed. The result is asecond order PDE, being used as approximation of blade flap-wisemotion equation. The form of this PDE is a nonlinear forced Mathieuequation, which describes the in-plane flap-wise vibration of fan blade.The frequency characteristic and mechanism of fan vibration are studied,based on which the simulation signal of fan vibration is presented andused in later analysis. The vibration signal is polluted by noise, thus anew improved EMD based denoising method is proposed to pre-processthe noisy signal.The feature of vibration signal is studied and extracted withsystematic mathematical method, the mathematical expression ofvibration signal features is presented and used in fan performance evaluation. The feature extraction operation of pre-processed vibrationsignal is as following:1. generate spectrum energy feature using ARmode based PSD.2. extract frequency component with improvedfrequency spectrum.3. compute statistical values in time domain as thetime domain feature. Construct the spectrum energy feature, frequencycomponent feature, time domain feature into a high dimension vector,which is seen as the feature representing the analyzed signal. Using PCAto merge and compress multi-index character.Fan performance evaluation with the approach of synthesizingaerodynamic performance and mechanical behavior isn’t systematicallystudied at home and aboard. This paper tries to do some initial analysisand research. With the basis on feature extraction of fan vibration signal,feature is merged and compressed based on methodology related topattern recognition, then the performance evaluation is realized in atheoretical analysis way.Using C#programming language, in accordance with therequirements of fan performance testing, fan performance testing andanalysis system is developed using serial port as communication mediabetween hardware and computer. The thought in software engineeringlike OOP and three-tier architecture is brought into practice in thissoftware, resulting in accurate and reliable testing and analysis of fanperformance with simple operations, in consideration of actual industrial practice, the automation of fan design is also studied and incorporatedinto the FPTAS.
Keywords/Search Tags:Modeling and analysis, EEMD-based denoising, Performance evaluation, System development
PDF Full Text Request
Related items