Hashemi, Seyyed Mahmoodreza2024-05-212024-05-212024-03Thesis 11363https://unbscholar.lib.unb.ca/handle/1882/37820In the active flow control, Stochastic Estimation (SE) is a beneficial tool to predict the time-resolved flow field as it offers acceptable short-in-line processing time with higher temporal and spatial resolution. In SE, a correlation between the desired variable and the corresponding time-resolved point measurements obtained from a few locations, must be built. These correlations are employed to estimate the desired flow property. This study delves into SE methods, comparing the conventional SE method with a relatively unexplored technique, Spectral Linear Stochastic Estimation (SLSE) in the flow over a backward-facing step. The study found that conventional SE underestimates fluctuations, suggesting improved accuracy with increased sensor points. SLSE, while enhancing estimation, tends to overpredict structures. A sensor placement analysis addressed these challenges, finding that 8-point sensors, concentrated before the reattachment point, yield the best accuracy. This research aims to advance the understanding and optimization of stochastic estimation for improved flow prediction.x, 97electronicenhttp://purl.org/coar/access_right/c_abf2Stochastic approximation.Flow visualization.Forecasting.master thesisHall, Joseph(OCoLC)1441256368Mechanical Engineering