In this work the data of the tire profile have been analyzed with different algorithms to identify tire faults and de-noise the signals. The four algorithms proposed are: Envelope, Erosion, Subset resampling and Wavelet. The four algorithms’ performances have been analyzed using simulated and experimental data acquired from a wheel balancer. In the simulations, the different factors influencing the measurement of the eccentricity components (modulus and phases) have been estimated with the four proposed algorithms; the influencing factors have been analyzed using the analysis of variance. Experiments were performed by acquiring the tire profile with noncontact sensors in different positions. Also in this case, results have been analyzed with the four proposed algorithms in order to extracts the eccentricity components. Results showed that the four algorithms were performing very well in de-noising the signal and can improve the measuring accuracy of run-out with respect to the simple DFT (the algorithm actually used on car wheel balancers with diagnostic capabilities).
Il presente lavoro di tesi descrive dei metodi numerici per l'identificazione del profilo degli pneumatici, al fine di separare la superficie di rotolamento dal disegno del battistrada. A valle dell'analisi della letteratura sono stati identificati 4 possibili metodi (Inviluppo, Erosion Filter, sottocampionamento locale e Wavelet denoising). Le performances dei 4 algoritmi sono state valutate tramite analisi numeriche e sperimentali. Nelle analisi numeriche sono stati simulati segnali differenti e l'effetto di diversi fattori (eccentricità, incertezza di misura, fill factor del battistrada) è stato analizzato mediante le tecniche dell'analisi della varianza. A valle delle simulazioni i test al vero sono stati effettuati su diverse categorie di pneumatici. I risultati hanno confermato quelli ottenuti nelle simulazioni: i diversi algoritmi permettono di migliorare l'efficienza rispetto al metodo attualmente utilizzato (DFT sul segnale grezzo).
Signal processing for the identification of tire faults
WU, XI
2013/2014
Abstract
In this work the data of the tire profile have been analyzed with different algorithms to identify tire faults and de-noise the signals. The four algorithms proposed are: Envelope, Erosion, Subset resampling and Wavelet. The four algorithms’ performances have been analyzed using simulated and experimental data acquired from a wheel balancer. In the simulations, the different factors influencing the measurement of the eccentricity components (modulus and phases) have been estimated with the four proposed algorithms; the influencing factors have been analyzed using the analysis of variance. Experiments were performed by acquiring the tire profile with noncontact sensors in different positions. Also in this case, results have been analyzed with the four proposed algorithms in order to extracts the eccentricity components. Results showed that the four algorithms were performing very well in de-noising the signal and can improve the measuring accuracy of run-out with respect to the simple DFT (the algorithm actually used on car wheel balancers with diagnostic capabilities).| File | Dimensione | Formato | |
|---|---|---|---|
|
2014_07_Wu.pdf
non accessibile
Descrizione: 2014-07-Wu Xi-Thesis
Dimensione
28.19 MB
Formato
Adobe PDF
|
28.19 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/93602