Real-time Neural Network based Identification of a Rotary-Wing UAV dynamics for autonomous flight

Real time flight implementation of a neural network based black-box identification (NNID) scheme to a rotary wing unmanned aerial vehicle (RUAV) is presented in this paper. The applicability of NNID scheme for real time identification of longitudinal and lateral dynamics of the RUAV is evaluated in flight. To show the efficacy of the method for real time applications, the identification results and error statistics are provided. The challenges involved in terms of hardware implementation, computational time requirements, and real time coding are investigated and reported. Results indicate that NNID is suitable for modeling the dynamics of the RUAV in real time.