Geometry simulation of stable channel bank profile using evolutionary PSO algorithm implementation in an ANFIS model

Forecasting the bank profile shape of stable hydraulic channels using empirical, experimental and numerical models is special consideration among fluid mechanic river science engineers. In present paper, application soft computing methods evaluated in predicting geometry channel cross sections. this way, a combination Particle Swarm Optimization (PSO) algorithm with an Adaptive Neuro-Fuzzy Inference System (ANFIS) model, novel evolutionary system called ANFIS-PSO presented. The model performance assessed comparison simple ANFIS model. coordinates points located on boundary state were also measured by authors sensor instrument laboratory at 4 different flow discharge rates 1.157, 2.18, 2.57 6.2 l/s. results indicate that Root Mean Squared Error (RMSE) Absolute Relative (MARE) 0.0132 0.1326 performed better than 0.026 0.1426 error values respectively (almost 97% 10% decrease RMSE MARE value for respectively). This demonstrates high accuracy characteristics. robust proposed can be used designing estimating dimensions. second-degree polynomial equation utilized other section.