Font Size: a A A

Analysis Of Smart Energy Optimization And Scheduling For Energy Internet System Based On Model Predictive Control Technologies

Posted on:2017-06-15Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y ZhangFull Text:PDF
GTID:1362330569498427Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
Energy internet system can locally and efficiently use the renewable energy resources,promote the user electricity quality and power supply reliability,it is one of the most important technologies of the future energy network.How to properly and effectively operate the energy internet is key to it.This dissertation proposes a general energy internet scheduling architecture which make the system stable,reliable and economical along the main line of optimization of local energy network,coordination of the multiple local energy networks and coorperation of energy internet users and utility companies.This scheduling architecture includes three layers: the local energy network layer,multiple local energy networks system layer and energy internet system layer,we then develop futher study of the architecture according to the different requirements in different layers from the aspects of time,space and function.The main contribution of this desertation are listed as follows.(1)A 'source-network-storage-load' integrated energy management model is set up which optimize the compresensive economic revenue of the local energy network and a standard model predictive control based local energy network optimal scheduling approach is proposed.The local energy network energy management model fully takes into account the special needs of energy storage system,distributed controllable resources,price inconsistence between the purchasing electricity price and selling electricity price in real-time electricity price market.Model predictive control framework is used to adjust the power operation of the energy storages and distributed energy resources timely to properly distribute the system power and decrease the negative impacts introduced by the forecast uncertainties of the renewable energy resources.(2)A stochastic model predictive control(SMPC)based local energy network optimal scheduling approach is proposed with considering the stochastic factos in the energy management model,it futher improve the robst performance of the scheduling approach to forecast uncertainties of the renewable energy resources.The typical uncertainty scenarios are chosen with scenario generation and scenario reduction methods,meanwhile,the two-stage based scenario reduction method proposed in this desertation can reduce the computation budern significantly to choose typical scenarios in large amount of initial scenarios,it promote the efficient of the scheduling approach.Moreover,the proposed SMPC based local energy network optimal scheduling approach consists of two parts: the pre-optimization scheduling stage and real-time power distribution and compensatation stage,it is more suitable for the system scheduling in energy internet condition with high penetration level of renewable energy resources.(3)A sequential distributed model predictive control(SDMPC)based mutilple local energy networks system optimal scheduling approach is devised by applying the noncooperative game theory.It not only can coordinate the operations of the local energy networks in the multiple local energy networks system where each local energy network has its individual objuective also reduce the negative impacts introduced by the randomness of the renewable energy resources.The energy router of each local energy network only communicate and exchange power with the energy router of the multiple local energy networks system,this mechanism decrease the communication and computational burdern of the distributed algorithm,reduce the risky of infringing each local energy network's privacy,and gurantee the safety of power supply.(4)A parallel distributed model predictive control(PDMPC)based energy internet system optimal scheduling approach is set up to coordination the operations of energy internet users and utility companies with the master-slave game theory.Due to the power demand amount of the whole nergy internet system is huge,the operation adjustement of the energy internet system must impacts the generation plan of the unility company and the electricity price in the electricity markt.Non-cooperative game models are studied among the energy internet user and utility companies,and between the user layer and utility company layer.
Keywords/Search Tags:Energy internet, Local energy network, Model predictive control, stoschatic programming, Game theory
PDF Full Text Request
Related items