A Novel Algorithm to Generate Synthetic Data for Continuous-State Stationary Stochastic Process (Wind Data Application)

摘要

Renewable energy resources have a great influence on country's future planning. However, building such investments needs reliable and accurate studies based on huge information and historical data, that might be considered a great challenge. Thus, generating some form of artificial or what they call it “synthetic” patterns that give the same hidden information and characteristics as the original records are so important. In this paper real wind speed data is gathered for a very promising candidate location. This dataset is used extensively to generate a synthetic data for planning/feasibility purposes for wind-farm project planning. Moreover, most commonly considered stochastic techniques were utilized to either model, extract all main probabilistic features of original data and hence; generate the required synthetic data. In addition, this paper proposed a new stochastic model that could generate synthetic wind data extremely has the same features as the original records.