New Concepts of Control Structure Design in Determination of PSS Location R. Asgharian A. Karimpour Dept. of Electrical Engineering, Mashhad University, Mashhad, Iran Abstract In a multi-machine power system, it is important to determine the best location for the application of power system stabilizers. A number of techniques have been proposed to perform this selection. In this paper a new selection measure based on relative gain array and singular value decomposition is proposed. A comparison is made between the performance of the new measure and the older methods. The proposed methodology is based on use of system transfer function. Key words--Power system stabilizer, relative gain array, singular value decomposition, controllability and observability measures. I Introduction Most available control design techniques assume that a control structure is given at the outset, but one of the first steps for an engineer is to determine the variables that should be controlled, the variables that should be measured, and the inputs that should be manipulated. An important task in the design of a control system is the specification of the control structure, referred to as control structure design. This leads to closing the gap between theory and application in this area [1-2]. Relative gain array (RGA) and singular value decomposition (SVD) [3], partial relative gains (PRG) [4] and closed loop interaction number (CIN) [5] have been introduced as measures of control structure design. In power systems, several techniques have been proposed in the literature to determine the best location for applying power system stabilizers [6-18]. Participation factor method is one of the methods in detecting the contribution of various generators in each mode and for suitable location of the PSSs [6]. Some authors have proposed sensitivity of PSS effect (SPE) as a criterion for determining the suitable location of PSSs [7]. Modal controllability and modal observability is also considered as a measure for determining the suitable location of PSSs and suitable signal as input of PSSs [8-9]. Relation among the most popular techniques for identifying suitable location for applying PSSs is given in [10]. For selection of best position for applying static var compensators, residue has been used as a screening index and final selection have done with relative gain array (RGA) [11]. RGA uses as a measure for selection of suitable location of stabilizers [12]. Singular value decomposition (SVD) has been proposed as a measure of the distance of a controllability and observability matrix from singularity in a state space model [8-9]. This distance is used as an index to compare the ability of inputs to control an oscillation mode. The main drawback of the above approaches is the determination of linear models from the state space models that usually have large dimensions. Some authors have used transfer