Showing papers by "Mohamed Tadjine published in 1997"

Invariant manifold approach for the stabilization of nonholonomic systems in chained form: application to a car-like mobile robot

Abstract: It is shown that the class of n-dimensional nonholonomic chained systems can be exponentially stabilized using the invariant manifold techniques. To this end, an invariant manifold, on which all the closed-loop trajectories tend to the origin, is derived for this class of systems under a linear smooth time-invariant state feedback. Thereafter, this manifold is made attractive by means of a discontinuous, time-invariant, state feedback. The proposed control scheme ensures exponential stability of the closed-loop system in a large domain excluding, at the initial time, the null value of the first state variable. Finally, our controller is applied for the stabilization of a car-like mobile robot and simulation results are given to highlight its effectiveness.

Discontinuous control design for the stabilization of nonholonomic systems in chained form using the backstepping approach

Abstract: This paper presents a new systematic way, based on the backstepping approach, to design discontinuous time-invariant state feedback controllers for exponential stabilization of nonholonomic systems in chained form. The enclosed simulation results show the effectiveness of the proposed controller.

Speed and rotor time constant estimation via MRAS strategy for induction motor drives

Abstract: This paper deals with simultaneous estimation of induction motor speed and its inverse rotor time constant for sensorless induction machine drive purposes. The proposed algorithm is based on the model reference adaptive system theory (MRAS). The simulation results seem to be very satisfactory for a large wide of supply frequency and torque load.

On the stability of an autonomous mobile robot subject to network induced delay

Abstract: We deal with the stability analysis of a mobile robot subject to network induced delay. First, we propose a discrete-time stabilizing controller in the delay-free case. Thereafter, the problem of induced random delay caused by the communication network is highlighted. Finally a predictor mechanism is developed to build a control law that stabilizes the mobile robot control system.

Stabilization of nonholonomic systems in chained form: application to a car-like mobile robot

Abstract: In this work the class of nonholonomic systems in chained form is studied. It is demonstrated that this category of strong nonlinear systems can be exponentially stabilized using an appropriate attractive manifold on which all the trajectories of the closed loop system tend to the origin, provided that the initial conditions are leading into an open and dense domain. To this end we propose a quasi-continuous state (output) feedback controllers for the n-dimensional systems in chained form. Simulation results are given to highlight the effectiveness of the proposed controllers.

A scheduling approach for decentralized mobile robot control system

Abstract: Most existing studies dealing with the scheduling of real-time control systems had neglected time variation in the perception, decision and actions loops of the controlled system, especially when this system is distributed over a broadcast network. In this paper, a distributed mobile robot architecture which takes into account the previous problem is proposed. To this end, a scheduling algorithm which allows one to determine accurately the message interarrival time at each node is developed. An application on a multiplexed field bus, named the controller area network (CAN), is also discussed.

Quasi-continuous output feedback control for nonholonomic systems in chained form

Abstract: In this paper, the problem of stabilizing the class of nonholonomic systems in chained form is solved by means of a quasi-continuous output feedback and invariant manifold technique. To this end, we assume that only the first and the last state variables are measurable. The proposed controller yields exponential stability in a large domain, while avoiding a discontinuity surface at the initial time. To highlight its effectiveness, the developed approach is applied to a nonholonomic wheeled mobile robot.

Adaptive sliding mode observer for state estimation of an induction motor on DSP-based system

Abstract: This paper presents a self-tuning variable structure observer to estimate the fluxes and the speed of an induction motor. Firstly, assuming that the speed is a known smooth bounded function, a sliding mode observer is designed to estimate the rotor fluxes. Then, using Lyapunov's technique and following the design approach of Kubota et al. (1993), an adaptation law is derived to estimate the unknown speed. Experimental results obtained on a digital signal processor (DSP) TMS320C31/40 MHz-based system are presented to highlight the effectiveness of the proposed technique.

Experimental evaluation of sliding mode induction motor control

Abstract: This paper presents a sliding mode approach for the control and states estimation of an induction motor. Firstly, the field oriented control is briefly reviewed. Secondly, a sliding mode control law is derived to regulate the speed and the rotor flux magnitude and a sliding mode observer which was proposed in [17] is used to estimate the unmeasurable part of the motor states. Finally, experimental results obtained on a digital signal processor (DSP) TMS320C31/40 Mhz based system for a 0.37 KW induction motor are given. The obtained results are compared with those obtained by the field oriented control system.