Multi-agent modeling and simulation of wind turbine behavior with failure propagation consideration

Abstract

Simulation is an interesting technique to represent and predict the behavior of complex system such as wind turbines. In order to get accurate representation of the real turbines, the main factors influencing the operating state of the system must be considered during the development of the model. Stochastic dependencies between components are among the important factors in determining reliability and unexpected system's failures. The main objective of our work is developing a multiagent simulator using the Netlogo platform that allows the representation of the behavior of a wind turbine and different failure scenarios considering the impact of stochastic inter-components dependencies, weather conditions and system configuration. A comparison between the reliability parameters is performed in both conditions (dependent and independent components). The results obtained show the usefulness of considering the existence of stochastic dependencies when predicting and optimizing the wind turbines reliability and availability.