Bi-level optimization of novel distribution network with VPP and flexible load cluster

Abstract

To satisfy the power requirements of modern industrial parks, while also enhancing energy efficiency and system economy, a bi-level optimization model is proposed for a novel distribution network, which incorporates a virtual power plant (VPP) and flexible load cluster. The proposed optimization model for the distribution network layer considers factors such as minimum voltage offset and operating cost. It accounts for uncertainties associated with new energy sources, control of the soft open point (SOP), as well as dispatching of the static var compensator (SVC). A normalized normal constraint (NNC) method is used to obtain Pareto front. The power flow model is converted into a normalized cone programming model utilizing second-order cone relaxation technique. The VPP optimization layer aims to achieve the lowest operational cost and comprehensively considers the complementarity of multiple types of distributed energy resources. The bi-level iterative solution is obtained using the analytical target cascading (ATC) method. Finally, the enhanced IEEE 33-node system example demonstrates that the proposed bi-level optimization model effectively reduces the system’s operating cost, optimizes the power flow distribution, and minimizes network losses.