Stochastic Petri Nets for very short-term wind speed modeling

To overcome the limitations of fossil energy and protect the environment from emissions of greenhouse gases, it is essential to develop the use of renewable energy as a substitute. At present, one of the renewable sources of energy is wind energy, which has the advantage of being pollution free and inexhaustible. However, the use of wind energy is strongly influenced by wind speed, which is not constant. Such varying wind speeds lead to the creation of fluctuated wind power. Consequently, there is a need for modeling and the accurate prediction of wind speed to help optimize the design of the turbine and control system in a wind energy conversion system to maintain system stability. This paper presents the modeling of very short-term wind speed using Stochastic Petri Nets (SPN) that is based on the measurement results of wind speed in Nganjuk. In this study, Stochastic Petri Nets was designed by using 7 places and 7 transitions. Transition to the SPN is defined as a function that generates random values using a uniform function. Wind speed data that was generated during a 500 seconds interval, was compared with the observed wind speeds. The comparison of the generated wind speed and observed ones shows that both its statistical characteristic have similar value.