Small low-cost unmanned aerial vehicle system identification: a survey and categorization

Small low-cost unmanned aerial vehicles (UAVs) provide greater possibilities for personal scientific research than other conventional platforms such as satellites or manned aircraft. In order to provide precision aerial imagery or other scientific data, an accurate model of vehicle dynamics is needed for controller development and tuning. The purpose of this paper is to provide a survey of current methods and applications of system identification (system ID) for small low-cost UAVs. This survey divides UAVs into 5 groups: helicopter, fixed-wing, multirotor, flapping-wing, and lighter-than-air. The current state of system ID research with respect to various types of UAVs is reviewed based on research literature. System ID methods and application are tabulated for further research. Concluding remarks are given and applications for system ID methods to small low-cost UAVs are recommended. Index Terms—System Identification; UAV, Helicopter, Fixedwing, Multirotor, Flapping-wing, Lighter-than-air, Least squares, Levenberg Marquardt, Kalman filter, EKF, UKF, Observer/Kalman identification, Autoregressive exogenous inputs, ARMAX, Box Jenkins, Prediction-error method, Output-error method, Neural network, Fuzzy identification, Time domain, Frequency domain, State-space, Subspace, CIFE