A sparse sampling approach to dynamic sub-cycle decomposition of apparent power in general polyphase networks

We present a dynamic decomposition of apparent power in a polyphase network, based on measurements taken within a portion of a single cycle of the ac fundamental. We show that two time samples suffice for a definition that provides an attractive trade-off between computational cost and dynamic performance. The sub-cycle (or near-instantaneous) metrics that we introduce connect the time- and frequency-domain approaches in a natural way, and address the key weaknesses of the existing instantaneous decompositions. Furthermore, our power metrics are amenable to refinements via symmetrical components, thus connecting with concepts commonly used in system protection. We illustrate the sub-cycle decomposition via two polyphase examples: one of conceptual interest, the other an actual industrial voltage sag event.