A nonparametric model in the task of predictive modeling of wind power plants

The work is devoted to solution of the problem of predictive modeling of relative capacity of wind power plants in relation to season and weather factors. It contains step-by-step description of the following steps of modeling: factor selection, raw data pretreatment, model evaluation and optimization. Both heuristic and formal methods were combined to construct the model. The basic modeling approach here is the k-nearest neighbors method. The model has been verified with the use of test sample. The developed model allows to optimize the wind power plant operation in relation to weather factors and network load.