Microgrid Operational Planning Using Deviation Clustering Within a DDDAS Framework

Abstract

As climate change progresses and the global population continues to increase, meeting the energy demand is an issue that has been brought to the forefront of the conversation. Microgrids (MGs) are groundbreaking tools that have risen in popularity to combat this crisis by capitalizing on renewable, distributed energy resources to efficiently satisfy the energy demand from environmental sensors via telemetry. In this work, we present a deviation clustering (DC) algorithm within a dynamic data-driven application systems (DDDAS) framework to reduce the length of the MG dispatch model’s planning horizon while retaining the temporal characteristics of the initial load profile. The DDDAS framework allows for the adjustment of the current dispatch decisions in near real-time. We develop two modules embedded within this framework; the first is a proposed rule-based policy (RBP) that modifies the sensing strategy and the second is the DC algorithm which reduces the execution time of the MG simulation. Numerical analysis was conducted on the IEEE-18 bus test network to assess the performance of the proposed framework and determine an appropriate threshold for clustering. The limitations of the presented framework were also determined by comparing the tradeoff between its the speed of the solver’s solution time and the accuracy of the resulting solution. The results indicate a decrease in solution time within the desired accuracy limits when using the proposed approach as opposed to traditional load dispatch.