In this paper we consider a typical problem of tracking of reference path variables for quadcopters. Due to the fact that the behavior of quadcopters is described by a multidimensional nonlinear system considering the influence of external disturbances, the known control algorithms are rather cumbersome and difficult to implement online. The classical methods of feedback linearization and disturbance suppression are deep feedback and sliding mode design methods, the implementation of which is often associated with physical infeasibility. In this paper, the above approaches are used to design state observers (which operate in a computational environment) and disturbances of a special kind, followed by feedback design in the form of combined control. The latter becomes possible due to the representation of the initial model of the control plant in the canonical space, which allows to provide the conditions of agreement on disturbances and model uncertainties.