Approximately Closed Loop Model Predictive Operation Control Of Grid-connected Microgrids

Micorgrids(MGs)which are composed of renewable energy generator,storage device,thermal generator and load are the next generation and the most important foundation of smart power system.Microgrids are a promising solution to global energy challenges and environmental pollution.One major question in this area is: how to operate a MG with high amount of renewable energy source(RES)in an optimal way,while providing load power and taking energy and power limits of the grid and the units into account.Robust and reliable operation while maximizing infeed from RES and reducing the use of thermal generation are important objectives in the control of MGs.In this thesis,three model predictive control strategies are applied to operate a grid-connected MG in the presence of bounded uncertainties in an optimal way,respectively,certainty-equivalence model predictive control(MPC),minimax MPC and approximately closed loop MPC.To find an optimal operation strategy,a mixed-integer quadratic program is solved online.In a numerical case study,the results of three approaches are analyzed.The simulation shows,that compared to the performance of the certainty-equivalent case,the minimax approach is robust to disturbances and allows a secure operation of MGs in grid-connected mode,but is more conservative with less RES and more thermal infeed;that approximately closed loop MPC can lead to an optimal operation control of MGs with less conservativeness,gurantee a secure and economic operation of MGs,higher amount of RES penetration and lower thermal infeed,can also successfully collaborate with the unitility grid,which indicate it is a reliable and practical energy management solution and can effectively improve the performance of the operation of grid connected MGs.