Non-Homogeneous Markov Wind Speed Time Series Model Considering Daily and Seasonal Variation Characteristics

Abstract

Wind speed model lays the foundation of wind power simulation and is crucial to the analysis of wind power integrated into power systems. This paper proposes a non-homogeneous Markov chain (NHMC) wind speed model that takes the daily and seasonal characteristics of wind speed variation into account. An optimal partition method is adopted to divide the wind speed time series into several segments affected by seasonal changes. A seasonal index is introduced before modeling to reduce the impact of seasonal variation. A time-related variable is also introduced to describe the daily periodic variation of wind speed. Evaluation on the probability distribution, autocorrelation function, and power spectral density of NHMC model and commonly used wind speed models is conducted. Moreover, the number of NHMC states on model performance is investigated. Simulation results demonstrate that the proposed approach offers excellent fitness on overall statistical properties of the wind speed time series.