A Cluster-based Aggregation of Shiftable Loads for Day-Ahead Scheduling

Abstract

The demand side flexibility offered by the residential smart appliances is being exploited to support the evolving concept of ‘demand-follows generation’. For a portfolio of a large number of loads, economic scheduling is a challenge due to individual load constraints and large problem dimensions. From this viewpoint, this paper provides an optimal day-ahead scheduling framework considering aggregated flexibility of a large number of residential loads. In the first step, cluster-based aggregation is performed to employ the flexible potential of shiftable load types which fully utilizes the detailed technical attributes of load data. In the second step, the aggregated attributes are utilized to optimally schedule the flexible load clusters subjected to the network constraints. A case study with 1000 shiftable loads is used to evaluate this strategy with and without clustering. It provides effective management of massive flexible loads to minimize the complexity of energy management and improve the overall economics with high computational efficacy. The results illustrate that this strategy can effectively reduce operational costs.