In this paper, a new design method for neural networks is presented ba
sed on evolutionary programming. By using an evolutionary algorithm th
e structure and weights of static neural networks can be simultaneousl
y acquired. This method is further extended to design recurrent neural
networks through introducing 'delayed links' into networks. Simulatio
n results are also given to illustrate the efficiency of the proposed
method. (C) 1997 Elsevier Science Ltd.