Neural networks based identification of helicopter dynamics using flight data

Identification of lateral and longitudinal dynamics of a helicopter using flight data is presented. Neural networks with linear filter also known as Narendra´s model and recurrent neural networks with internal memory (memory neuron networks) are used for the purpose. These networks are used for identifying the rate and attitude response of the helicopter to a given longitudinal cyclic and lateral cyclic inputs. The relative effectiveness of these algorithms has been discussed.