Predicting Discharge in Tidally Affected Rivers Using Genetic Algorithm

摘要: In the present work, new expressions for discharge in tidally affected rivers have been derived employing genetic algorithm (GA). The proposed expressions use few geometric and hydraulic characteristics of rivers which are readily available to field engineers. The developed GA models were applied in forecasting flows of Nakdong River at Samrangjin Gauging station in South Korea, which is highly affected by tidal waves. The comparative study of the prediction by the proposed expressions and by those obtained from the robust minimum covariance determinant method (RMCD) suggests superiority of new expressions for predicting flows in tidally affected rivers.