Predictive Model Based Energy Scheduling And Optimization For Islanded Microgrid Systems

Due to the intermittent renewable energy,diverse types of loads,microgrids bring challenges in energy management.Therefore,in order to achieve active distribution network security,stable and economical operation,we need to do novelty research on islanded microgrids energy management.Based on predictive model,this thesis studied some energy management problem related to bidirectional disturbance in both supply and demand sides.The main research works and contributions of thesis are summarized as follows:1.Due to uncertainties of renewable sources,improve the renewable energy utilization of distributed microgrid systems,this paper presents an optimal predictive model based graded energy strategy to coordinate energy management among islanded microgrid systems.In particular,through information exchange among systems,each microgrid in the network,which includes renewable generation,storage systems and some controllable loads,can not only maintain its own system-wide supply and demand balance,but also reduce the system operation cost.With our mechanism,the closed-loop stability of the distributed microgrid systems can be guaranteed.In addition,we provide evaluation criteria of renewable energy utilization to validate our proposed method.Simulations show that the supply-demand balance in each microgrid is achieved while,at the same time,which demonstrates the effectiveness and efficiency of our proposed policy.2.With the increased penetration of Renewable Energy Sources(RESs)and plug-andplay loads,islanded Micro Grids(MGs)bring direct challenge in energy management due to the uncertainties in both supply and demand sides.Based on model predictive control,two energy scheduling strategies are proposed.The first method is Battery Energy Storage Systems(BESSs)as the main scheduling unit,which optimizes the charge/discharge energy in order to satisfy the load demand.Moreover,the enhanced DMPC strategy is designed to guarantee the RESs efficiency and necessary demand response.The second one is Distribution Network Operator(DNO)as the main scheduling unit.A two-stage control scheme based on MPC is proposed for coordinating energy scheduling and optimizing Demand Response(DR)of Autonomous MicroGrids(AMGs)when there is an imbalance between the RES supply and demand.Firstly,DNO will make an optimal decision based on the information received,which include trading with the other AMGs,using the energy from BESS,or adjusting the controllable loads.Secondly,the optimized operational options will be tracked by MGs over a long time horizon.3.Due to randomness of charging/discharging and uncertainties of energy dispatch,we present a coordinated energy dispatch based on DMPC,where the upper level provides an optimal scheduling for energy exchange between DNO and microgrids while the lower level guarantees a satisfactory tracking between supply and demand.With the proposed scheme,not only we maintain a supply-demand balance in an economic way,but also improve the renewable energy utilization of distributed microgrid systems.To describe the dynamic process of energy trading,a novel conditional probability distribution model is introduced,which can characterize randomness of charging/discharging and uncertainties of energy dispatch.Moreover,we formulate a two-layer optimization problem and the corresponding algorithm is given.Finally,simulation results show the effectiveness of the proposed method.4.Internal demand may exceed internal power supply provided by RESs and BESSs in AMGs.To derive a balance for the mismatched demand response and energy supply,a three level hierarchical coordination strategy is proposed.The top level is responsible for energy coordination between the DNO and AMG.The DNO will purchase/sell energy from/to an AMG that has surplus/deficient energy at a slow sampling period.The medium level focuses on the local balance of each individual AMG,which optimizes the charge/discharge energy of BESSs and dispatches of the aggregator with the same sampling period as the top level.The bottom level will make load cutting decisions according to the optimization results of the medium level in the case of supply-demand imbalance,which is updated of a fast sampling rate.Furthermore,the two-time scale optimization scheme is applied to reduce the effects of bidirectional disturbances caused by the randomness of RES operation and elastic loads,as well as efficiently solve a different time scale energy scheduling.Simulation results show the effectiveness of the proposed methodology.