!

DISSIPATIVE ANALYSIS AND SYNTHESIS OF TIME-DELAY SYSTEMS

!
M. S. Mahmoud, H. N. Nounou
In this paper, complete results are established for the problems of dissipative analysis and control synthesis of a class of linear time-delay systems. The developed results encompass all available results on approach, passivity and positive realness for time-delay systems as special cases. Both delay-independent and delay-dependent cases are investigated and all the sufficient stability conditions are cast as linear matrix inequality-based feasibility tests. For control synthesis, state feedback is considered. Then robust dissipativity results are derived for the case of convex polytopic uncertainties. Numerical examples are provided.

SOLUTION OF CERTAIN CLASSES OF LINEAR TIME VARYING SECOND ORDER SYSTEMS

!
K. Busawon, P. Johnson
In this paper, we derive the analytical solution for certain classes of linear time varying systems described by second order linear ordinary differential equations. Since it is well known that a second order linear homogeneous differential equation can always be transformed into a Riccati equation, we first derive the solution of some particular classes of Riccati equations. As a result, the solution of a fairly large class of homogeneous second order linear systems can be obtained. Next, the methodology for providing the solution of certain classes of second order linear inhomogeneous systems is given. We show, in particular, that solving a second order inhomogeneous linear differential equation is equivalent to solving a Riccati equation and a first order linear equation that are coupled together. In addition, some examples are provided to demonstrate the procedure employed.

ON VECTOR CONTROL OF DOUBLE STAR SYNCHRONOUS MACHINE WITH CURRENT FED INVERTERS

!
L. Nezli, M. O. Mahmoudi, M. S. Boucherit, M. Djemai
This paper deals with the study of vector control of the double star weak salient synchronous machine. The machine is supplied with two current inverters controlled via hysteresis based PWM strategy. The main objective is to achieve a decoupling between rotor flux and the electromagnetic torque, by introducing a vector control which will maintain the stator current producer of the electromagnetic torque in quadrature with rotor flux. Moreover, to optimise the machine torque as in the case of a DC machine, we have introduced a tuning of the excitation current through a chopper at chopping frequency of 5 KHz feeding the rotor excitation circuit. Simulation were carried to demonstrate the efficiency of the proposed approach.

SOFTWARE AGENTS IN THE CONTROL AND SURVEILLANCE OF A HYDROELECTRIC POWER PLANT CHAIN

!
 L. Miclea, Sz. Enyedi, H. Valean, I. Stoian
Intelligent software agents are software modules that able to make decisions on their own, communicate with each other, learn new things and travel from one system to another. This article presents a modern control project that uses software agents to further improve the performance of a real time control and monitoring system of a complex heterogeneous system. The agent system will be superimposed over the existing supervisory control and data acquisition system (SCADA) of multiple hydroelectric power plants connected in a chain on a river, aims to improve performance in terms of efficiency, stability, and fault tolerance.

PROFICIENCY TESTING PROGRAM UNDER NABL IN THE PRESSURE RANGE 7- 70 MPa USING A DEAD WEIGHT TESTER

!
S. Yadav, A. K. Bandyopadhyay
This paper describes the proficiency testing (PT) of the seven laboratories, accredited by National Accreditation Board for Testing and Calibration of Laboratories (NABL), India having best measurement capabilities better than 0.05% of full-scale pressure by means of interlaboratory comparison using dead weight tester (DWT) as an artifact in the pressure range 7 – 70 MPa. The program started in November 2002 and completed during May 2004. National Physical Laboratory (NPLI), New Delhi has the privilege of coordinating this programme and also acted as a reference laboratory. The comparison was carried out at 10 arbitrarily chosen pressure points (7, 10, 15, 20, 25, 30, 40 50, 60 and 70 MPa) throughout the entire pressure range of 7 – 70 MPa. The normalized error (En) value of six laboratories out of the total seven are found < 1in the entire pressure range and are acceptable well within their reported expanded uncertainty at a coverage factor k = 2. These results are quite encouraging for the manufacturers and calibration laboratories that are facing a new environment after the WTO agreement.

FUSION OF DELAYED MEASUREMENTS AS A LEAST-SQUARES OPTIMIZATION PROBLEM: AN APPLICATION TO PLANETARY EXPLORATION

!
A. Brandes, S. Monaco
This paper deals with fusion of delayed measurements into an existing state estimate, formulated in terms of a non-quadratic least squares optimization problem. Two approaches are considered to enable a closed-form analytical solution by means of an approximated quadratic cost function: a standard one and a newly proposed approach. Error analysis and a simulation study revealed better adequacy of the new approach for the problem at hand. A case study regarding planetary exploration is presented and discussed. Simulation results confronting the approximate analytical solution with standard quasi-Newton optimization procedures demonstrate the effectiveness of the proposed approach.

ADAPTIVE OUTPUT TRACKING OF A CLASS OF NONLINEAR SYSTEMS IN POLYNOMIAL LOWER-TRIANGULAR FORM

!
B. Wang, H. Ji, Z. Chen, H. Xi
This paper investigates the problem of adaptive output tracking for a wide class of nonlinear systems, called polynomial lower-triangular form (pLTF), which expands LTF to a more general case. Aiming at the structure of system, the adaptive controllers are explicitly constructed to ensure the global stability property and practical output tracking, by adding and modifying a power integrator. The pLTF relaxes the restrictions on the structure of the normal LTF and broadens the family of systems. Finally, the examples show the practical usage of this class of systems and the effectiveness of the design approach.

ANALYSIS OF GAIN MARGIN AND PERFORMANCE OF INVERSE OPTIMAL CONTROLLERS

!
A. Shahmansoorian, A. Khaki Sedigh, B. Moshiri, S. Mohammadi
Optimal stabilization guarantees several desirable properties for the closed loop system, including stability margins. In a direct approach the Hamilton-Jacobi-Belman (HJB) equation has to be solved, which in general is not a feasible task. This motivated both Freeman and Kokotovic to pursue the developments to design methods that solve the inverse problem of optimal stabilization. In this paper, the stability margin and conservativeness properties of the inverse optimal controllers have been presented. It has been shown that the use of Control Lyapunov functions (CLFs) obtained from feedback linearization, in Sontag’s formula generally generates a conservative suboptimal solution for an optimization problem and the achieved control law does not posses proper gain margin. A class of inverse optimal controllers is developed, which has low gain margin and low control effort. A performance criterion for Sontag’s formula has also been presented.

A DISTRIBUTED WIRELESS SENSOR NETWORK LOCALIZATION SCHEME MOTIVATED BY GRAPH RIGIDITY THEORY

!
S. Wang, K. Liu, F. Hu, J. He
In this paper, we focus our work on a new distributed nodes localization algorithm while the distribution of the nodes is in random fashion. Some recent research results on the graph rigidity theory are introduced and incorporated. Moreover, we present a practical scheme for node localization. The major contribution of this paper is that we have constructed the localizable collaborative body to model network localization and applied it to test conditions for unique localizability. Simulation results showed the effectiveness of the proposed localization scheme in terms of localization accuracy and scalability.

NECESSARY CONDITIONS FOR DISTRIBUTED CONTROL PROBLEMS GOVERNED BY PARABOLIC VARIATIONAL INEQUALITIES WITH AN INFINITE NUMBER OF VARIABLES

!
S. El-Zahaby, G. Mostafa
In this paper, the necessary conditions for optimality in distributed control problems governed by parabolic variational inequalities with an infinite number of variables are established. The optimality conditions are expressed in terms of generalized gradients, and are obtained by means of an abstract approximating control process. Using theorems of Barbu and  pervious results of Gali et al and El-Zahaby.