Online Multiperiod Power Dispatch With Renewable Uncertainty and Storage: A Two-Parameter Homotopy-Enhanced Methodology

In this paper, an online multiperiod power dispatch with renewable uncertainty and storage is presented. The ac power flow equations, voltage limits, and thermal limits are respected to ensure the obtained solution causes no static violation while ensuring network transfer capability. A scenario reduction technique using affinity propagation clustering is developed to reduce the number of scenarios and relieve the computational burden. A three-stage solution methodology is developed to solve the nonconvex constrained optimization problem, in which a two-parameter homotopy-enhanced methodology with a decomposition scheme is developed to reliably and quickly solve the online multiperiod power dispatch problem. The proposed three-stage solution methodology has been evaluated on several testing systems, ranging from a 30-bus system to a 3012-bus system with four time periods, and the numerical results are promising.