Alejandra Ferreira

Robust Control With Exact Uncertainties Compensation: With or Without Chattering?

The problem of robust exact output control for linear systems with smooth bounded matched unknown inputs is considered. The higher order sliding mode observers provide both theoretically exact observation and unknown input identification. In this paper, a methodology is proposed to select the most adequate output control strategy for matched perturbations compensation. The aim of this paper is to investigate the possibility for exact uncertainties compensation using signals identified by high order sliding mode observers. Towards this aim, we modify the hierarchical super-twisting observer in order to have the best observation and identification accuracy possible. Then, two controllers are compared. The first one is an integral sliding mode controller based on the observed values of the state variables. The other strategy is based on the direct compensation of matched perturbations using their identified values. The performance of both controllers is estimated in terms of the deterministic noise upper bounds, sampling step and execution time. Based on these estimations, the designer may select the proper controller for the system. Experimental results are given for an inverted rotary pendulum system.

Sliding mode identification and control for linear uncertain stochastic systems

This paper presents the integral sliding mode technique applied to identifying disturbances and robustifying the optimal linear quadratic Gaussian (LQG) controller for linear uncertain stochastic systems, which is compared to the conventional sliding mode approach. The obtained identifier/controller provides a method for estimating uncertainty values and ensures robustness of the system against perturbations throughout the entire response starting from the initial time instant. Numerical simulations illustrating the obtained results are given for the inverted pendulum system

Robust regulation for a 3-DOF Helicopter via sliding-modes control and observation techniques

In this paper, two robust control strategies for a 3-DOF Helicopter via sliding-modes techniques are presented. First, quasi-continuous controllers along with a sliding mode differentiator are designed and then the design of classical PID controllers in combination with a second-order sliding mode observer is presented. Both strategies preserve high position regulation accuracy and robustness to model uncertainties and external disturbances. Simulations and experimental results on a 3-DOF Helicopter by Quanser are shown.

Output hierarchical super twisting observation based robust feedback control: With or without chattering?

The problem of robust output feedback control for linear systems with unknown inputs based on hierarchical super twisting (HST) observation and identification is considered. Theoretically, the finite time convergence of HST observation and identification errors allows the compensation of the matched uncertainties. The accuracy of the observation, the identification, and the compensation is discussed in terms of upper bounds for deterministic noise, sampling step, and execution time. The LQ control for the nominal system is considered as a case study. Two methods to be compared are proposed: compensation via the identification of the uncertainties, which allows to avoid the chattering, and compensation via integral sliding mode control. Based on this comparison, the designer could select the more adequate controller for the system. Numerical simulations and experimental results are given for an inverted rotary pendulum system.

Higher Order Sliding Mode observers for actuator faults Diagnosis in robot manipulators

A diagnostic scheme for actuator faults which can occur on a robot manipulator using a model-based Fault Diagnosis (FD) technique is addressed. With the proposed FD scheme it is possible to detect a fault, which can occur on a specific component of the system. To detect actuator faults, higher order sliding mode Unknown Input Observers (UIO) are proposed to make analytical redundancy. The observers input laws are designed according to the so-called Super-Twisting Second Order Sliding Mode Control (SOSMC) approach and they are proved to be capable of guaranteeing the exponential convergence of the fault estimate to the actual fault signal. The proposed approach is verified in simulation and experimentally on a COMAU SMART3-S2 robot manipulator.

LQG-robust sliding mode control for linear stochastic systems with uncertainties

This paper presents the integral sliding mode technique robustifying the optimal linear quadratic Gaussian (LQG) controller for linear stochastic systems with uncertainties, which is compared to the conventional sliding mode approach. The obtained controller ensures robustness of the system against perturbations throughout the entire response starting from the initial time instant and provides a method for evaluating uncertainty values. Numerical simulations are given for the inverted pendulum system

Output nested backward compensation of unmatched effects of unknown inputs

This paper tackles the regulation problem of linear time invariant systems with unmatched perturbations. The proposed methodology exploits a high order sliding mode observer, which guarantees theoretically exact state and perturbation estimation. In this work is introduced a controller with a compensation strategy based on the identified perturbation values. A nested backward sliding surface design is proposed, which allows to compensate the unmatched uncertainties and to stabilize some of the non-actuated state components, while all the remaining states are maintained bounded. The overall stabilization error is estimated in terms of the sampling time and actuator time constant values. The feasibility of the technique is showed in a rotary inverted pendulum simulation example.

Sliding Mode Identification and Control for Linear Uncertain Stochastic Systems

This paper presents the integral sliding mode technique applied to identifying disturbances and robustifying the optimal linear quadratic Gaussian (LQG) controller for linear uncertain stochastic systems, which is compared to the conventional sliding mode approach. The obtained identifier/controller provides a method for estimating uncertainty values and ensures robustness of the system against perturbations throughout the entire response starting from the initial time instant. Numerical simulations illustrating the obtained results are given for the inverted pendulum system.

Identification based generation of self-excited oscillations for underactuated mechanical systems via two-relay algorithm

A tool for generating a self-excited oscillations for underactuated mechanical systems by means of a variable structure controller is proposed. The design procedure allows to find the explicit expressions of the controller gain parameters in terms of the desired frequency and amplitude through describing function (DF) method. Necessary conditions for orbital asymptotic stability of linear systems are derived via a modification of the Loeb criteria. The identification and compensation method yields an exact linearization of the original nonlinear system avoiding imperfections on oscillations characteristics. Performance issue of a self-excited system, applied to a Furuta pendulum, is illustrated by simulations.

Nested backward compensation of unmatched effects of unknown inputs based on HOSM identification

The paper considers the regulation problem of linear time invariant systems with unmatched perturbations. The proposed methodology exploits a high order sliding mode observer, which guarantees theoretically exact state and per- turbation estimation. It is introduced a controller with a compensation strategy based on the identified perturbation values. When the system satisfies quite restrictive assumptions, the method ensures exact regulation of the unmatched states. In order to deal with the general case it is proposed a nested backward strategy to design the sliding surface, which allows to compensate the unmatched uncertainties and to stabilize some of the non-actuated state components, while all the remaining states are maintained bounded.