Review of adaptive decomposition-based data preprocessing for renewable generation rich power system applications

Time series decomposition is extensively used recently for time series forecasting. The obtained set of relevant monocomponents using an adaptive decomposition method are potential candidates for predictions using point and probabilistic forecasting frameworks. Time series decomposition has been widely applied to the volatile time series of input variables of various power system analyses. This paper comprehensively represents computational steps for different adaptive decomposition methods and enlists open issues for each family of methods. The attributes of the decomposition method are reviewed in detail. Furthermore, the metrics for their characterization are formulated, and the significances are highlighted. The applications of time series decomposition to various data preprocessing activities are discussed. Finally, the solutions adopted in the literature in selecting appropriate parameter(s)/function(s) associated with decomposition methods, suppressing undesirable effects due to a method, and solutions to reduce the computational complexity in decomposition are summarized. This comprehensive review is expected to provide a clearer picture of existing decomposition methods and research scopes for a novice reader in time series decomposition.