Steady-state self-oscillations and chaotic behavior of a controlled electromechanical device by using the first Lyapunov value and the Melnikov theory

•Analysis of an Andronov–Poincaré–Hopf bifurcation as a function of a control law.•Analysis and simulation of weak focus stability by using first Lyapunov value.•Investigation of the homoclinic orbits and the Melnikov method to predict chaos.•Complete analysis of chaos indicators from a computational viewpoint.•Analysis of control laws by using chaotic behavior in the presence of random noise.

In this paper regular and chaotic oscillations in a controlled electromechanical transducer are investigated. The nonlinear control laws are defined by an electric tension excitation and an external force applied to the mobile piece of the transducer. The paper shows that an Andronov–Poincaré–Hopf bifurcation appears as long as adequate parameters are chosen for the nonlinear control laws. The stability of the weak focuses associated to such bifurcation is examined according to the sign of the first Lyapunov value, showing that chaotic behavior can arise when the first Lyapunov value is varied harmonically. The appearance of a homoclinic orbit is investigated assuming an approximated model for the device. On the basis of the parametric equations of the homoclinic orbit and the presence of harmonic disturbances on the platform, it is demonstrated that chaotic oscillations can also appear, and they have been examined by means of the Melnikov theory. Chaotic behavior is corroborated by means of the sensitive dependence, Lyapunov exponents and power spectral density, and it is applied to drive the transducer mobile piece to a predefined set point assuming that noise due to the measurement process can appear in the control signals. The steady-state error associated to such random noise is eliminated by adding a PI linear controller to the control force. Numerical simulations are used to corroborate the analytical results.