Open Channel Junction Velocity Prediction by Gene Expression Programming and Regression Methods

Open Channel Junction Velocity Prediction by Gene Expression Programming and Regression Methods

مینو شریفی پور1 حسین بنکداری2 امیر حسین زاجی3

1) دانشجوی کارشناسی ارشد عمران گرایش سازه های هیدرولیکی، دانشگاه رازی کرمانشاه -
2) دانشیار گروه مهندسی عمران - دانشگاه رازی کرمانشاه -
3) دانشجوی دکتری عمران گرایش سازه های هیدرولیکی، دانشگاه رازی کرمانشاه -

کد IOI مقاله : XYCE-ACFFZ
محل انتشار : کنفرانس بین المللی عمران ، معماری و زیرساخت های شهری(icica.ir)
چکیده :
    Open channel junctions are one of the most significant and practical structures in hydraulic engineering. Due to erosion in the contraction zone and sediment deposition in the separation zone, the flow velocity in the junctions is a vital topic in the designing of open channel junctions. This study aims at an accurate prediction of the flow velocity in open channel junctions through Gene Expression Programming (GEP) and regression models, by use of different points of the flow (x*, y*, and z*) and the ratio of the upstream to downstream discharge in the main channel (q*) as input parameters. The numerical models are compared according to their different statistical errors. It is concluded that the GEP model with Mean Square Error (MSE) value of (0.055) is more accurate in predicting longitudinal velocity in open channel junctions than the regression models with MSE values of 0.103.

کلمات کلیدی : open channel junction velocity gene expression programming regression methods