Research On Multi-time-scale Rolling Optimization Scheduling Of Source-load-storage

With the increasingly serious problems of energy crisis and environmental protection,clean,high-efficiency renewable energy represented by wind power and photovoltaics has attracted great attention and has been widely used in power systems.However,due to the wind speed and light intensity are intermittent,fluctuating,and random due to environmental influences,there are inevitably errors and uncertainties in forecasting output power.Therefore,under the background of vigorously developing renewable energy,the total amount of uncertainty in the system is also increasing with the increase in penetration rate.Microgrid,as an important form of fusion of renewable energy and large power grid,will cause a lot of disturbances to the economic and reliable operation of microgrid due to a large amount of uncertainty.In order to make better use of renewable energy and ensure the economical,efficient,reliable and stable operation of the microgrid,this paper starts from solving the disturbance of the system caused by the error and uncertainty of source and charge prediction to study the economic optimal scheduling of the microgrid.The research of this paper mainly includes the following aspects:1)Firstly,this paper discusses the research status of microgrid in various countries under the background of energy crisis.A microgrid system including wind power generation,photovoltaic power generation,diesel engine and energy storage is taken as the research object,and the related mathematical model is established.2)Secondly,this paper proposes a multi-time-scale optimization scheduling method.In the day-ahead long-time-scale,based on wind,photovoltaic output and load demand power of the cumulative probability distribution curves,using Latin Hypercube Sampling(LHS)to discrete continuous probability distribution model to obtain samples,then obtaining the typical scenarios by synchronizing cutting back,and the selection probability of the biggest scenes as wind,photovoltaic output and load demand power forecast economic optimization,adopting the Particle Swarm algorithm(PSO)model to obtain the economic optimization scheduling plan.In the intraday short-time-scale,using Model Predictive Control(MPC)to track the optimal operation plan as the goal on the basis of actual running state and roll-revise the schedule plan in real-time.3)Thirdly,aiming at improving the shortcoming of stochastic optimization,this paper proposes the robust optimization in the day-ahead stage to deal with the lowfrequency component of source and load prediction errors and uncertainties,and establishes the robust optimization model considering the demand responses with the goal of economic optimization,which makes the day-ahead scheduling plan can adapt the intraday actual running state of wide fluctuations,and uses the PSO to solve the model to obtain the robust economic optimization scheduling plan.In the intraday stage,MPC is adopted to deal with the high-frequency component of source and load prediction errors and uncertainties,and also is used to roll-revise the schedule plan in real-time.4)Finally,this paper summarizes the research results,and prospects the future research directions of multi-time-scale rolling optimization.