Active Optimal Dispatch of Microgrid Based on Improved Model Predictive Control

Abstract

In order to cope with the uncertainty of intermittent distributed power supply output in the microgrid, and to solve the problem that the prediction accuracy is declining with the increase of time scale, this paper uses deep learning technology to achieve accurate prediction of new energy output. Based on this, an optimal dispatch of micro-grid based on model predictive control is proposed to fine-tune the coordination and control of wind power, photovoltaic and energy storage equipment in the microgrid so as to maximize the dissipation of the intermittent distributed power supply and track the microgrid operation reference trajectories accurately.