Ecology, Environment and Conservation Paper

APPLICATION OF GENETIC ALGORITHM IN OPTIMIZATION OF DIMENSIONLESS PARAMETERS TO THE ARTIFICIAL NEURAL NETWORK MODEL FOR PREDICTING ALLUVIAL RIVER SUSPENDED LOAD TRANSPORT

Mona Omidvarinia and Mahmood Shafai Bajestan

Abstract

Estimation of the suspended load of rivers are very important factors in river training, water supply projects and reservoir sedimentation. Therefore, it is important to use the new techniques which estimates the sediment load as accurate as possible. In this regard, artificial intelligent techniques such as Artificial Neural Network and Genetic Algorithm have been proposed as alternative to conventional techniques to solve a wide range of problems in various domains. Once a neural network model has been created, it is frequently desirable to use the model backwards and identify sets of input variables, which result in a desired output value. The large numbers of variables and non-linear nature of many materials and the algorithm of train models can make finding optimal solutions difficult and often fall in the local minimal. Genetic algorithms (GAs) are search algorithms based on the mechanics of natural selection and genetics as observed in the biological world. Importantly, to implement a genetic algorithm for training of the Artificial Neural Network this problem would recover. The propose of this research is to use the Artificial Neural Network and Genetic Algorithm for estimation of the sediment load in Karoon river at two stations of Molasani and Ahwaz.(was used as a case study that located in Khuzestan in southwest of Iran). The genetic Algorithm was applied to train the Artificial Neural Network and the results compared with Artificial Neural Network trained with back-propagation algorithm. In this research Input of network were dimensionless parameters, where 5 cases were defined to run the model. On the other hand, the simulation result (for all 5 cases) have been analyzed and compared with experimental method and sediment rating curve in order to show the importance of this methodology.