A variable dead time circuit breaker auto-reclosing scheme using artificial neural networks

This paper presents an advanced methodology of an auto-reclosing relay for the transmission line protection in a power system. The failure of high-spe ed auto-reclosing of a circuit breaker can result in electrical or mechanic al stress on the power system equipment and apparatus. When a power system has a sufficient degree of transient stability, the proposed scheme can giv e rise to a higher rate of successful reclosing by increasing the dead time of a reclosing relay. This paper presents a method of artificial neural ne tworks (ANN) based on-line quantitative measurement of the relative degree of power system transient stability in terms of the stability margin derive d from the transient energy function method. The significance of this metho d is that it uses only the information obtainable easily in the local area around the fault location instead of system-wide information in order to de termine the stability margin by virtue of ANN for the on-line application t o the control of auto-reclosing dead time. (C) 1999 Elsevier Science Ltd. A ll rights reserved.