Research On Energy Management Strategies Applied To Energy Router

Energy router,based on power electronics technology,is an intelligent device that integrates the management and control of all power generation,electricity consumption and energy storage equipment in energy local area network(LAN).As an indispensable component unit in the future energy Internet,energy router is of great significance in the aspects of maximum consumption of renewable energy,intelligent power consumption on the user side,stable operation of power grid,etc.Based on the above research background,this paper takes the energy router and its local area network as the research object,and aims to improve the utilization rate of renewable energy and save the electricity purchase cost of users to study the energy management strategy oriented to the energy router.Firstly,this paper builds the model of the energy LAN equipment where the energy router is located.Then,as the basis of energy management,short-term and super-short-term forecasting of traditional electricity load is carried out.In view of short-term load forecasting,the input data scheme is determined by the control variable method,and the support vector machine regression forecasting algorithm is improved to reduce the error of short-term load forecasting.For the super-short-term load forecasting,the input data scheme is also determined by the control variable method.On the basis of short-term load forecasting,the rolling forecasting is carried out,and then the super short-term load forecasting with high forecasting accuracy is obtained.Secondly,as an emerging power in the energy Internet,the scale of use of electric vehicles is growing.The charging load of electric vehicle is studied in this paper.The research is divided into two parts:one is the disordered charging of electric vehicles.The Monte Carlo method is used to simulate the random charging of electric vehicles,and different load curves in multiple scenes are obtained.Through comparison,it is proved that the charging curves in more scenes are more consistent with the actual situation.Second is orderly charging of electric vehicles.In order to reduce the adverse impact of large-scale disordered charging of electric vehicles on energy Internet,an orderly charging method of electric vehicles based on residual SOC was proposed.Through the comparison of simulation results,the effectiveness of the proposed method is proved.Finally,the multi-time scale energy management strategy is studied by using electricity load prediction and EV orderly charging load as input.In the case of grid connection,economy and user satisfaction are taken as the goals to conduct day-ahead energy management.Then,in order to deal with the influence of uncertainty such as prediction error,a new real-time energy management strategy based on day-ahead energy management is proposed,which adjusts the day-ahead optimization curve in real time by updating the input.By comparison,the multi-time scale energy management strategy can well deal with the influence of uncertainty under the condition of global optimization when grid-connected.For the case of isolated island,the multi-time scale energy management strategy of planned isolated island is similar to that of grid connection,except that the objective function is to reduce the cutting load,which is also verified by simulation in this paper.For the real-time management strategy of unplanned islands,this paper proposes a scheduling method based on decision tree,which is proved by simulation to be a good solution to unplanned islands.At the end of this paper,Matlab and LabVIEW were used to jointly build the energy management platform.