Optimal Operation Of Active Distribution System Based On Source Network Load Storage

Because of the gradual shortage of fossil energy and environment problem which cannot be ignored,at the same time,the electricity demand of society and industry is increasing,renewable energy generation has made great development in recent years,particularly wind power and photovoltaic.The proportion of distributed power generation is increasing and high proportion renewable energy integrated into distribution system puts forward higher requirements for its planning and operation.The technology of active distribution system emphasizes the flexibility and initiative of the system and exerts active control on schedulable resources in order to reduce the impact of renewable energy volatility.At last,the efficient fusion of source,network,load and storage is realized.Firstly,this paper carries out detailed analysis and modeling on various kinds of resources in active distribution system,selectively analyzes the uncertainty mechanism of renewable energy output and demand side resources,then establishes its probability model.Secondly,the paper introduces the concept of scenario analysis theory,scenario generation and reduction,then proposes a multi scenario generation method considering correlation between scenes and a scene reduction method based on clustering technology.Thirdly,a coordinated control strategy based on multi time scale for active distribution system is proposed,then establishes optimal scheduling models of three time-scales.In day-ahead dispatch,the prediction error is large,so the dispatch result is used as an input for in-day dispatch.The in-day model corrects the day-ahead dispatch result caused by prediction error step by step.As a result,the utilization of renewable energy and the optimal operation capability of the system are improved,the cost of system operation is reduced.The model also takes system security constraints into consideration.For the proposed nonlinear multi-objective optimization model with many constraints,the improved multi-objective particle swarm optimization is proposed for model solving.In the algorithm,the speed of the particle random perturbation strategy and adaptive inertia weight strategy is designed to improve the convergence speed and global optimization capability.Finally,the modified IEEE33 node distribution system is selected as an example.The results show that the multi time scale optimization and uncertainty model can accommodate renewable energy output fluctuation and promote dissipation of renewable energy.The correlation of scene reduces the deviations between scenes as well as the operating cost of the system.