Constraints in Rainfall-Runoff Modeling Using Artificial Neural Network

Abstract

This chapter discusses constraints in prediction of Inflow to reservoir using Multi Layer Perceptron-Artificial Neural Network (MLP-ANN) Technique. Rainfall-runoff (R-R) information is required to provide basic information for reservoir management in a multipurpose water system optimization framework. The relationship between rainfall and subsequent inflow to reservoir (R-R) is an extremely complex and difficult problem involving many variables, which are interconnected in a very complicated way. ANNs are flexible mathematical structures that are capable of identifying complex non-linear relationships between input and output data sets. A neural network consists of a large number of simple processing elements that are variously called neurons, units, cells, or nodes. There are no specific rules for determination of number of hidden layers and hidden neurons in each layer, selection of learning rule, activation function, training criteria and input parameters, as well. Activation functions are needed for introducing non linearity into the network and it is non linearity that makes multilayer networks so powerful.