The model of Nano quad-rotors contains many uncertainties such as an external disturbance from a wind field, highly non-linear strong coupling between variables and body measurement errors. To deal with these uncertainties and control the Nano quad-rotors, a novel data-based disturbance observer (DO) is firstly proposed to observe disturbances from a wind field and perturbations from errors of parameter estimation. Then the DO is used to improve the conventional dynamic inversion (DI) method to obtain an enhanced dynamic inversion (EDI) method, which relies only on roughly estimated geometrical parameters, thus eliminating the largest flaw of conventional DI, namely depending on detailed plant information. Simulation results show that the method proposed achieved good trajectory tracking with only roughly estimated geometrical values under wind field; the DO proposed can accurately estimate disturbance from a wind field and perturbation from error of parameter estimation.
