Data mining method for battery operation optimization in photovoltaics

Recently, a photovoltaic (PV) system has attracted attention because of serious environmental and energy problems. In near future, PV systems intensively connected to the grid will bring about the difficulties in the power system operation. As a countermeasure, this paper deals with the introduction of storage battery for making the unstable PV power controllable. In this regard, when we introduce a storage battery into a PV system, we have to consider the advantages and disadvantages. In order to evaluate the system from various perspectives, we have carried out multi-objective optimization of battery operation in PV system design. However, as the number of objective functions increases, it becomes difficult to appropriately interpret the correlation among objective functions and design variables. With this background, in this paper, a novel computational method is proposed for data mining of PV system design, in which we make an attempt to effectively extract the design information of the battery system with the use of Self-Organizing Map (SOM).