Research On Energy Management Framework For Intelligent Electricity Distribution And Utilization Systems In Islanded And Grid-connected Modes

As a new generation of power network,the smart grid utilizes the advanced sensing,measurement,communication,computer,control and other technologies to achieve the comprehensive goals of flexible and efficient,safe and reliable,green and economic power generation,transmission,distribution,and utilization.As an important part of the strong smart grid,the intelligent electricity distribution and utilization systems(IEDUSs)are facing with many challenges brought by the limitation of energy resources,the uncertainty of renewable energy generations(REGs),the complexity of multi-energy flow coupling,and the security of private information in the process of energy development,while developing effective energy management and optimal dispatching strategy is the key point to ensure the economic and efficient operation of the IEDUSs.In view of this,this dissertation carries out research on energy management and optimal scheduling of the IEDUSs in the isolated operation mode and the grid-connected operation mode,respectively.The main contents are listed as follows:For a special kind of the energy management and optimal scheduling problem of a pelagic islanded microgrid in the isolated operation mode,this dissertation proposes a longperiodic energy trading framework.Part of the limited energy resources in the pelagic islanded microgrid comes from the local intermittent REGs,while the other part is transported fuels and storage batteries periodically by ships from the main land.The two types of end users(strategic users and normal users)in the pelagic islanded microgrid have different priorities of energy supply,where the strategic users have higher priority of energy supply than normal users.Considering the variability and randomness of the weather on the pelagic islands,it is extremely difficult to obtain the probability distribution of the REGs,thus this dissertation adopts the robust optimization approach to address the high uncertainty associated with the REGs brought by the long period,designs a decentralized bi-level iterative algorithm to solve the non-convex optimization problem derived from the gametheoretic approach and backward induction,which effectively protects the users' privacy information,and finally obtains a Stackelberg equilibrium strategy to maximize the revenue of aggregator and minimize the energy cost of each user.The effectiveness and advantages of the proposed periodic energy trading framework are verified by numerical simulations.Compared with a baseline periodic energy trading system and a short-periodic energy trading system,the simulation results show that the energy demand of strategic users can be preferentially satisfied in the cases of sufficient and insufficient energy supply under the decentralized periodic energy trading system,thus it is more suitable for the pelagic islands.For a special kind of the multi-energy management and optimal scheduling problem of the pelagic islanded microgrid clusters(PIMGCs)in the isolated operation mode,this dissertation designs a multi-energy management framework for a PIMGC.Part of the limited energy resources of the load islands comes from the local intermittent REGs,while the other part is transported electricity and natural gas energy resources by ships from the resource islands.This dissertation proposes a hierarchical multi-scale energy management strategy,where the operators on resource islands determine their daily optimal energy supply in a distributed collaborative way based on the primal-dual subgradient method;the aggregator on each load island determines its daily optimal energy demand and hourly optimal energy usage;the user on each load island determines its hourly optimal energy consumption.This dissertation proposes a tri-level day-ahead distributed iterative algorithm to protect the users' privacy information,and finally obtains a Nash equilibrium strategy,where the operators minimize their aggregate operational cost,each aggregator maximizes its revenue and each user maximizes its payoff.The numerical simulation tests are carried out to demonstrate the effectiveness and advantages of the proposed multi-energy management framework.This dissertation conducts comparison simulations from three aspects of whether the energy supply is sufficient,different distances between resource islands and load islands,different robustness parameters of the REGs,the simulation results show that the multi-energy demand of users on load islands can be satisfied as much as possible in the cases of sufficient and insufficient energy supply under the multi-energy management framework,and it can effectively tackle the variability and uncertainty in the PIMGCs.For a more general kind of the energy management and optimal scheduling problem of the IEDUSs in the grid-connected operation mode,this dissertation proposes a real-time demand response framework.The real-time two-way interactions are conducted between a utility company and multiple residential users equipped with energy storage systems in a IEDUS.This dissertation establishes a temporally-spatially coupled convex optimization model.Based on the dual decomposition method,the temporally-spatially coupled convex optimization problem is decoupled into several independent subproblems.This dissertation designs a distributed real-time algorithm,the residential users and utility company solve the subproblems in a distributed way to find the optimal energy management scheduling scheme for each user and utility company,which not only maximizes the overall social welfare,but also effectively protects the privacy information of users.Compared with no demand response strategy and real-time demand response strategy without considering energy storage system,the simulation results show that the proposed real-time demand response strategy can not only decrease the peak energy demand of users and the generation cost of utility company,but further save users' power expenses and increase the social welfare.