Control of a twin-rotor mechanism in conditions of incomplete information

In the paper we consider the twin rotor mechanical system as a control plant. The behavior of this system is similar in many ways to the behavior of a real technical device — a helicopter. The tracking problem of the generalized coordinates (yaw and pitch angles) for the given signals is posed under conditions of parametric uncertainty and the action of uncontrolled disturbances. In the synthesis of the control law, the decomposition principle and nonlinear $\mathrm{S}$-shaped feedbacks in the form of sigma functions are used. Under the assumption that only angular position sensors are available, the reduced order state observer with sigmoidal corrective actions is constructed for estimating angular velocities. It is shown that the use of sigma functions in feedbacks provides: transient responses of the controlled variables that are close to aperiodic; tracking for the given signals which is $\varepsilon$-invariant with respect to the existing uncertainties without expanding the state space due to internal models; the ability to take into account design restrictions on velocity and control at the synthesis stage. The presented results of computer simulation confirm the effectiveness of the developed algorithms.