OPTIMUM DESIGN OF AEROSPACE STRUCTURAL COMPONENTS USING NEURAL NETWORKS

The application of artificial neural networks to capture structural de sign expertise is demonstrated. The principal advantage of a trained n eural network is that it requires a trivial computational effort to pr oduce an acceptable new design. For the class of problems addressed, t he development of a conventional expert system would be extremely diff icult. In the present effort, a structural optimization code with mult iple nonlinear programming algorithms and an artificial neural network code NETS were used. A set of optimum designs for a ring and two airc raft wings for static and dynamic constraints were generated using the optimization codes. The optimum design data were processed to obtain input and output pairs, which were used to develop a trained artificia l neural network using the code NETS. Optimum designs for new design c onditions were predicted using the trained network. Neural net predict ion of optimum designs was found to be satisfactory for the majority o f the output design parameters. However, results from the present stud y indicate that caution must be exercised to ensure that all design va riables are within selected error-bounds.