Design and optimization of model predictive control (MPC) for energy efficient microgrid

This research focuses on the DC microgrid system combined with the photovoltaic (PV) arrays and the control mechanics for maximum power point tracking (MPPT) as its object. The main problem tackled is that power extraction from PV systems is very inefficient due to variation of the environment or load that conventional MPPT approaches cannot effectively handle. The superior MPPT performance of this study’s novel dual mode model predictive control (MPC) approach is derived from a new dual mode model predictive control (MPC) approach. The implemented system shows RMSE of 7.0085 for conventional MPPT methods, whereas tracking efficiency is maintained within between 94.8 % to 97.2 % of the maximum power, which is available. Under standard test conditions, the system achieved less than 0.15 sec response time and less than 0.45 sec settling time, while degrading less, yet handling various environmental changes. It is due to the MPC’s predictive capability and real time optimization framework. The major contributions of the proposed solution, among others, include its dual mode design that supports both left and the right side regions in the PV curve together with the integrated charging management of the battery, as well as its robust constraint handling that enables safe operation and maximal power extraction. This system is well suited to implementation in small to medium scale DC microgrids that can be tolerant of up to 800 W/m2 per second of irradiance variations and up to 50 °C temperature range and demonstrated several hundred kilocycles hours of stability. The solution can offer practical benefits to the grid connected and standalone PV systems with the requirements of rapid response to environmental change and high extraction efficiency of power