Abstract of Thesis presented at COPPE/UFRJ as a partial fulfillment of the requirements for the degree of Master of Science (M.Sc.)

Uncertain Systems Control Based on the Super-Twisting Algorithm for Global Exact Tracking

Paulo Victor Nunes Monteiro Vidal

October/2016

Advisors:

Eduardo Vieira Leão Nunes Liu Hsu

Department:

Eletrical Engineering

      In this dissertation, super-twisting control techniques are designed for uncertain systems in the presence of nonlinear state-dependent uncertainties/disturbances. A new design for a recently proposed multivariable super-twisting algorithm is introduced, allowing to deal with multivariable systems with uncertain symmetric positive definite input matrix. Using a Lyapunov function approach, properties of global finite-time convergence for this new design are demonstrated. In addition, motivated by a satellite motion stabilization problem, a nonlinear state-dependent time scaling is used in order to obtain a variable gains algorithm which is able to present global finite-time convergence properties in the presence of a broader class of disturbances. Another important contribution of this dissertation is to propose monovariable and multivariable output-feedback control strategies based on the Variable Gain Super-Twisting Algorithm. To this end, a multivariable non-decoupled extension for this algorithm is introduced. In order to obtain an output-feedback solution, a norm bound for the unmeasured state variable is estimated using first order approximation filters. Global uniform asymptotic stability and finite-time convergence properties are proved by means of a Lyapunov function approach. Simulation results illustrate the effectiveness of the control strategies proposed in this dissertation for systems of practical interest.