In this thesis, artificial neural networks (ANNs) are used to model helicopter main rotor dynamic loads as a function of flight variables. The motivation to develop an accurate model of such loads is to reduce maintenance and replacement costs by eliminating excessive conservatism currently associated with structural fatigue estimation. Neural networks are used for the modeling procedure because of their capability to model complex, nonlinear relationships for multiple input-multiple output systems.

In support of the dynamic loads modeling discussed above, this thesis also briefly reviews artificial neural network technology, and investigates the modeling of a well-known dynamic system using ANNs.