Research On Robust Scheduling Of Distributed Generation In Stand-alone Microgrid

With the deterioration of the global energy crisis,search for unconventional energy has become an urgent demand.While the fragility of the ecological environment has put forward strict requirements for the development and utilization of new energy resources,the transformation of energy structure promotes the development of new power system.The use of new energy sources to promote the birth of Distributed Generation Technology.With the development of distributed generation technology,microgrid emerges as the times require.Microgrid is made up for the lack of new energy base on distributed energy,which includes many kinds of distributed power supply,energy storage system,load,energy conversion system and protection management unit.Economic operation and energy management are the key technology of microgrid.In the practical application of microgrid,many factors need to be considered,including the structure and composition of microgrid,the uncertain factors of distributed power supply and the uncertainty and fluctuation of load.This paper is based on the microgrid of wind turbine,photovoltaic cell,fuel cell,micro gas turbine and diesel generator.It regards safety,economy and environmental protection as a requirement.According to the randomness,intermittence and fluctuation of the output and load of the photovoltaic wind power in time and space,the optimal operation of microgrid must change from the deterministic optimization problem to the incomplete or uncertain optimization problem,and the research on robust scheduling of distributed generation in microgrid has become a hot and difficult issue in the research field:(1)Receding horizon optimization scheduling:first,build a micro-grid optimization scheduling model by combining dynamic programming and short-term prediction,in the given scheduling window,in order to minimize the total cost under scheduling window the and the cost of micro grid operation environment as the goal,to the safe operation condition and the relationship between the constraints of scheduling period.Next,the cell membrane optimization algorithm is used to solve the example,and the convergence speed and convergence of the algorithm are compared with the differential evolution algorithm.Last,the influence of the scheduling window on the rolling optimization scheduling of the microgrid is analyzed through experiments.(2)Robust optimization scheduling:first,according to the fluctuation of wind power output and load,an improved scheduling model based on robust optimization theory is proposed.This model constructed a robust optimal scheduling model for microgrid with uncertain predictive control strategy,which makes the minimum of the operation cost,environmental cost and micro power fluctuation as a target,considering safe operation and spinning reserve requirements as constraints.Next,an improved multi-objective differential evolution algorithm is used to solve the problem.Then,analyze the influence of uncertain budget on the micro grid scheduling,and compares the performance of the scheduling with the traditional scheduling.And in view of the fact that the conservatism of the robust optimal scheduling will cause the energy storage system to be over charged and discharged,and the uncertainty of the uncertain budget is difficult to be determined,the paper puts forward the adjustment strategy of the microgrid uncertainty prediction.Last,as the result of multi objective scheduling is an optimal solution set or an optimal target set,the grey entropy correlation degree is taken as the evaluation index.(3)Chance constrained programming scheduling:in view of the uncertainty factors in micro grid,the chance constraint is used to improve the microgrid scheduling model to realize the soft constraints of microgrid scheduling.The model is based on the minimum cost of operation and environmental cost,and the constraints of safe operating conditions and reserve requirements are used to construct a chance constrained programming model for the microgrid with event distribution adjustment strategy.Based on the probability distribution of the wind power prediction error,the probability distribution of the PV prediction and the load probability distribution,the distribution function of the event is constructed to generate the multi scene knowledge set.Finally,Monte Carlo simulation is used to compare the confidence level of the traditional chance constrained scheduling model.