The Hierarchical And Stochastic Optimal Dispatch Of Virtual Power Plant Containing Micro-Grid

Virtual power plant is the collection of multiple distributed generations, loads and storage. It can manage large-scale distributed generations which have different ownership and located dispersed. It can also integrate several micro-grids in different areas, manage them by its communications technology and present them to the main grid unified, lighten their burden on the main grid. The core of virtual power plant is its energy management system, which can coordinate energy distribution between generations, loads and storage. Because the diversity and particularity of distributed generations’ types and control modes, the energy management and optimal operation of virtual power plant has more complexity. The optimal dispatch of virtual power plant containing micro-grids not only needs to consider the coordinated operations of micro-grids, but also needs to achieve the optimal dispatch themselves. A hierarchical and stochastic optimal method to virtual power plant containing micro-grids is presented in this paper.(1)The types, characteristics and mathematical models of distributed generations in virtual power plant is presented, the Copula theory and its application in intermittent distributed generations is expounded and the stochastic and correlations of wind and photovoltaic generations is analyzed in this paper. The joint probability density models of the outputs of several intermittent distributed generations based on Copula theory is established.(2) Based on the theory of hierarchical optimization and chance constrained stochastic optimization, a hierarchical and stochastic optimal dispatch model of virtual power plant is established in this paper. The upper layer is the virtual power plant layer, aiming at the overall profit maximization of virtual power plant and considering the constraints of generation plans, the lower layers are the micro-grids layers, aiming at the cost minimization of micro-grids and considering the constraints of energy balance, the outputs constrains of distributed generations. The micro-grids layer also consider the stochastic and correlations of intermittent distributed generations such as wind and photovoltaic generations, whose optimal models are all chance constrained optimal models.(3) The sample average approximation conversion method of chance constrained model (SAA) and the use of optimization software GAMS is researched in this paper. Then the influences of intermittent distributed generations’ correlations, SAA sampling frequency and confidence levels on the results of stochastic optimization is analyzed. The numerical example results shows the effectiveness of the proposed virtual power plant hierarchical and stochastic optimal dispatch method.