Decreasing of dependency to expert in dynamic risk analysis along with machine learning methods

Risk management is a crucial tool for facilities for managing decision makings and operations. Also, risk management is essential for energy sector because of recent energy supply crisis around the world. Risk analysis which is a key tool for risk management has been carried out by the way of different techniques until today. As in many fields, machine learning has been used for risk analysis along with technological developments. It is aimed that decreasing dependency to expert in this study associated with using of machine learning techniques for risk analysis. So, a SCADA related database which consists of operational and alarm data of a wind turbine is examined and system failures and economical losses-based risk models are investigated. Features which are effects risk levels are included to risk model automatically and it is aimed to decrease expert opinions. At the end of the study, either real-time based or future-based risk predictions are observed with high consistencies.