Use of Genetic Algorithm in Development and Optimization of Nonlinear Dynamic Rainfall-Runoff-Sediment Yield Prediction Model

摘要

A nonlinear dynamic rainfall-runoff-sediment yield prediction model was developed for Kashinagar watershed of Vamsadhara river basin, Orissa (India). Model order was determined by multiple correlation analysis and then based on 't-test'. A promising and relatively new approach, that is, Genetic Algorithm has been employed for the estimation of model parameters and function optimization. The model-simulated sediment yields were in the close agreement with observed sediment yields.