Design of Adaptive Structure using Neural Network.

Abstract

This paper describes an application of a hierarchical neural network (HNN) to design technique of an adaptive structure. HNN is used to estimate the adaptive design solutions of a statically determinate truss structures (SDTS) subjected to variable load conditions. The learning data of HNN, i. e., the beam length of SDTS and the load condition, are calculated using the 9eneralized reduced gradient method and thd maximum stiffness law. The effectiveness of HNN for the design of adaptive structures is presented for several numerical examples.