In this article, the advantages and the key issues of the genetic
algorithm evolved neural network has been presented to model the
rainfall–runoff relationship in Ourika catchment. Our methodology
adopts a real coded GA strategy and hybrid with back-propagation
algorithm. The genetic operators are carefully designed to optimize
the neural network, avoiding problems premature convergence
and permutation. The experiment with real rainfall–runoff data
have showed that the predictive performance of the proposed
model is better than that of the traditional BP neural network. This
has been supported by the analysis of the changes of connection
weights and biases of the neural network.
One problem when considering the combination of neural network
and genetic algorithm for rainfall–runoff forecasting is the
determination of the optimal neural network topology. Our neural
network topology described in this experiment is determined manually.
A substitute method is to apply the genetic algorithm for
neural network structure optimization, which will be a part of
our future work.