Reducing power system models by hankel norm approximation technique

Abstract

This paper introduces a simplified algorithmic approach for reduction of power systems models using Hankel norm approximation technique. Mathematical models of physical systems derived from theoretical considerations are usually complex and of high order. Such systems are difficult to be analysed, so it is desirable to reduce such models to a low-order model. The algorithm discussed for computation of balancing transform is based on orthogonal transformation which extends the applicability of algorithm to non-minimal models. This method can be applied for order reduction of any linear time invariant, minimal/non-minimal or continuous/discretetime systems. For illustration of proposed algorithm, examples of power system models are considered. Resulting reduced order models are guaranteed to preserve the time domain and frequency domain characteristics. Further, it is shown that proposed technique gives better H∞ performance error than existing techniques.