Bayesian Modeling of Road Surface Roughness Characteristics for Fatigue Damage Assessment

Please enable javascript in your browser.

Mobassrfar, Yasin | 2023

113 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 55981 (45)
University: Sharif University of Technology
Department: Aerospace Engineering
Advisor(s): Adibnazari, Saeed; Shariyat, Mohammad
Abstract:
Road roughness-induced vibrations are the main source of fatigue damage accumulation in vehicles. Hence, road surface profile modeling and simulation are of high importance when it comes to vehicle fatigue damage assessment. This research focuses on uncovering the statistical distributions that describe the characteristics of road loads that affect fatigue damage accumulation. Particle filtering is deployed to estimate the log-volatility of the road surface profile by assuming a random walk behavior for the hidden log-volatility. The skew-normal distribution with three parameters is fitted to the estimated log-volatility. The inferred parameters are used for synthesizing artificial road profiles used in a vehicle simulation to calculate fatigue damage indices (FDI). The new road profile model outperforms all the previous models in predicting the real road FDI.
Keywords:
Stochastic Volatility Model ; Moving Average (MA) ; Particle Filter ; Bayesian Method ; Bayesian Inference ; Skew-Normal Distribution ; Fatigue Damage Index (FDI) ; Road Surface Roughness

No TOC