Hierarchical Coordination of Heterogeneous Flexible Loads

This paper first proposes a simple yet effective load model to characterize a variety of flexible loads. We then present a hierarchical control framework to coordinate their energy consumption. The upper level of the framework optimizes the aggregate power trajectory of the loads via a two-stage stochastic program, which aims to minimize the total energy costs under uncertain real time electricity prices and renewable generation. At the lower level, we design priority-list based scheduling algorithm for each load type to allocate the resulting power trajectory to individual loads. We then present several numerical examples to compare the performance of the framework against three benchmarks including a fully centralized, globally optimal approach. We find that, while the centralized framework is computationally intractable at even modest numbers of loads, the hierarchical framework scales very well to large numbers of loads and results in electricity costs that are comparable to the centralized benchmark.