Energy Management Strategy Of Multi-Microgrids System Based On Distributed Model Predictive Control

The energy management system of microgrids can provide a dispatching plan to ensure basic supply and demand balance for microgrids according to the forecasting information.However,in the islanded multi-microgrids system with high penetration of renewable energy,the accuracy of forecasting information will be greatly affected by the high degree of uncertainty and volatility of renewable energy.And the great challenges will be brought to the accuracy and robustness of the scheduling plan,which will lead to problems of the stability and security operation of the microgrids.Therefore,in order to deal with the problems caused by the uncertainty of renewable energy,this thesis conducts a series of studies based on model predictive control under different operating modes of islanded multi-microgrids system.Firstly,considering the influence of uncertainty in the multi-microgrids system which each microgrid operates independently,an energy management strategy based on model predictive control is proposed.The proposed strategy uses the idea of rolling optimization to form an intra-day scheduling strategy,which can reduce the burden of optimization calculation.On this basis,a feedback module is designed to form a closed-loop feedback structure,in which the ultra-short-term prediction and the output correction module can effectively cope with the uncertainty of renewable energy,and ensure the security of the operation of the microgrid.Secondly,considering a multi-microgrids system with multi-operaters coexistence,in which they are connected with each other,a distributed model predictive control strategy is proposed.An auction power market based on non-cooperative game is designed to maximize the benefits of individual microgrids.Meanwhile,the distributed strategy is used to minimize the amount of information communication among microgrids,to ensure the privacy of the operation information of individual microgrids.Then,to further improve the robustness of scheduling strategy,an energy management strategy based on robust model predictive control is proposed.This strategy combines the uncertain set model of robust optimization with the intra-day rolling framework of model predictive control;the robustness and economy of the scheduling strategy are balanced effectively,and also the effective operation of the scheduling plan within the uncertainty range will be ensured.In addition,the improved feedback prediction model is used to improve the accuracy of the feedback module which can improve the ability of the microgrid to cope with the fluctuations of renewable energy under the independent operation mode of the multimicrogrids system.Finally,in order to further consider the fairness of the electricity trading market and increase the participation enthusiasm of the members,a distributed robust model predictive control-based energy management strategy is proposed.An incentive electricity trading market based on alliance is designed,and different alliance strategies are designed for different participants.The seller takes the power contribution as the index to carry out the competition game,while the buyer uses the priority factor to achieve the fair distribution of benefits.The cost of microgrid group is reduced based on the characteristics of the distributed strategy to minimize the required information communication.At the same time,the real data and typical islanded multi-microgrids system structure are used to simulate all the proposed strategies,and the feasibility and advantages of the strategies are analyzed.