Model predictive control of Microgrid based on Mixed Logic Dynamic

Abstract

By analyzing the hybrid property of Microgrid in detail, a mixed logical dynamical model is set up with Hybrid theory. The energy in DC bus capacitor is chosen as state variable, the system functions and constraints with continuous state variables, input control variables, logical variables and auxiliary variables are presented. Based on the MLD model, model predictive control theory is applied for the purpose of reducing the batteries switching times between charge and discharge, the DC bus voltage fluctuation is compressed simultaneously. The quadric form objective function is designed and optimal input control variable sequence is achieved by mixed integer quadratic programming method. A simulation prototype is established, simulation results testify the proposed method can both suppress DC bus voltage fluctuation and decrease switching times of battery energy storage system. Index Terms -Microgrid; Hybrid theory; mixed logical dynamical model; model predictive control