Clusterization for Distributed Timely Detection of Changes in Smart Grids

Abstract

For real-time management of power distribution systems, when rapid operational adjustments are required to cope with intermittent power generation which is typical of renewable-based units, it is imperative that the optimization of the overall power distribution system be addressed in a distributed fashion. Then, the power distribution system may be partitioned into clusters whose size is determined by the delay constraints induced by the regular operations and the required operational adjustments. In this paper, clusterization is considered as directly addressing the operational adjustment problem in the presence of operational changes. Then, such changes need to be identified timely and accurately before pertinent adjustments be performed. Clusterization may thus be dictated by the accuracy and delay constraints imposed on the detection and identification of such changes. In particular, we first consider the initially non-clusterized power distribution network and determine the bus voltage and/or current variations perceived as considerable changes.  Then, we formulate a recursive maximum likelihood (ML) approach which naturally points to an initial network clusterization via incorporated sufficient identifiability conditions. We subsequently develop, analyze and evaluate a distributed sequential detection of change algorithm, implemented by the supporting data communication network, whose performance (including accuracy and decision delay) is controlled by a set of threshold parameters.  Required algorithmic performance constraints may dictate final cluster architecture and dimensionality. This performance monitoring and clusterization approach has never been considered in power systems before.