An MPC-Based Dual-Solver Optimization Method for DC Microgrids With Simultaneous Consideration of Operation Cost and Power Loss

Abstract

In this paper, a dual-solver framework based on model predictive control (MPC) is proposed, $E-$ solver and $L-$ solver. The economic scheduling problem is formulated using mixed-integer linear programming (MILP), which can be solved in an efficient way by using commercial solver ( $E-$ solver). While the transmission loss problem is formulated using non-linear programming (NLP), which can be solved in the interior point method, namely $L-$ solver. The $E-$ solver provides an economic priority power scheduling plan for the $L-$ solver, and the $L-$ solver solves the entire microgrid accurate power flow scheduling plan. The proposed planning model decomposition technique aims to solve the planning model in a time-sharing manner and combines the characteristics of the two optimizers with a reasonable matching algorithm to achieve economic, efficient, and fast real-time control. A case study of a DC microgrid is employed to assess the performance of the online optimization-based control strategy. Simulations based on eight-node DC microgrid show that the method reduces the operating cost by $12.10\%$ and increases the calculation speed by $80.18\%$ compared with the traditional interior point method. With a shorter delay, the proposed optimization method will facilitate the implementation of real-time control and optimization.