Performance analysis of different predictive models for condition monitoring of direct drive wind turbine generator

This paper presents a performance analysis of predictive models for the generator module which can be used as reference improvement in condition monitoring system using wind turbines farm Taiwan. With being critical component prone to failures, it is important perform data on its parameters that could monitoring. The main innovative feature this framework conduct before development system. Also, consistency between different evaluated. are generated neural network algorithm with combination from SCADA correlation well mean square error were then computed analysis. Results showed pairing input higher output parameter would give better model. Furthermore, was consistent throughout indicates same model developed and belonging farm. Employing preliminary combinations help optimizing