Research On Energy Management Strategies Applied To Community Power Router

With the increasing energy consumption of the community,renewable energy has gradually penetrated into the energy supply to reduce the fossil energy shortage and environmental pollution.Developing smart community that can effectively consume renewable energy and improve the utilization of renewable energy generation has become an inevitable requirement for energy structural transformation.The community power router(PR)provides access ports with communication and power conversion functions for renewable energy generation equipment,energy storage systems(ESS),and user loads,serving as the hub of information and energy.It is of great significance to ensure the stable and economic operation of smart community to coordinate renewable energy generation and user electricity consumption and achieve intelligent management and scheduling of energy.Based on the above research background,this paper takes community PR as the research object and studies the energy management strategy of the community PR,with the goal of improving the utilization of renewable energy,reducing electricity costs,and minimizing the impact of renewable energy generation and user consumption fluctuations on the grid.Firstly,this paper establishes a smart community energy distribution network architecture considering the PR.As the core device of the smart community,the community PR is connected to all the generating and consuming devices for information and energy interaction.Then,the characteristics of renewable energy generation equipment,ESS and user loads are analyzed,and corresponding mathematical models are established,which provide the foundation for subsequent energy management strategy.In addition,based on the smart community energy distribution network architecture constructed in this paper,the corresponding PR operation mechanism is proposed.The price-based demand response mechanism is used on the energy-consuming side,and the multi-time scale energy management strategy is adopted on the energy supply side,which provides the underlying framework for the energy management strategy of the community PR.Secondly,to improve the utilization of renewable energy generation,the photovoltaic power and wind power in the smart community are predicted.To address the problem of traditional attention mechanisms being unable to distinguish the importance of input variables,a long short-term memory(LSTM)network based on dual-attention mechanisms is proposed for photovoltaic and wind power prediction.The network obtains weights for different time steps through self-attention mechanisms to extract key information in the time dimension.Then,the attention mechanism is used to obtain weights for the row vectors of the hidden layer state to extract key information in the variable dimension.The prediction result of the current time step is obtained by combining the hidden layer vector of the current time step and the context vector of the time and variable dimensions.Simulation results show that the proposed LSTM based on dual-attention mechanisms can effectively extract key information from input data and further improve the accuracy of renewable energy generation prediction.Additionally,this paper proposes a multi-time scale energy management strategy for community PR based on PR operation mechanism proposed in this paper,including day-ahead energy management(DAEM)strategy and real-time energy management(RTEM)strategy.In the process of formulating the DAEM strategy,a Stackelberg game model between the PR and users is established.The PR indirectly adjusts the electricity consumption strategy of users through the community internal electricity price,and users adjust their electricity consumption strategy to affect the community internal electricity price.Users only report their total electricity load in each period without disclosing specific details which protects the privacy of users.Based on the DAEM strategy,the RTEM strategy is proposed to deal with the fluctuations of renewable energy generation and user load,which is divided into grid-connected RTEM strategy and islanded RTEM strategy.The former aims to reduce the impact of energy fluctuations on the grid,while the latter aims to ensure the basic electricity demand of users.They respectively adjust the power of the ESS and schedule the energy within the community.Experimental results show that the proposed multi-time scale energy management strategy can effectively reduce electricity costs and minimize the impact of energy fluctuations on the grid in grid-connected mode.In islanded mode,this strategy can ensure the basic electricity needs of users,and maintain high energy storage capacity to cope with emergencies.Finally,this paper designs a software platform for the energy management system of the community PR,and analyzes the main functions of the energy management system as well as the framework structure and the device composition of the PR.The platform is programmed with MATLAB and LABVIEW,and the renewable energy generation prediction model and the multi-time scale energy management strategy proposed in this paper are applied to realize the functions of the renewable energy generation prediction,day-ahead and real-time energy management,state monitoring and control of the PR energy management system.