Intelligent Optimisation of a Wind Turbine Digital Twin Model

Abstract:
In this work a digital twin framework is proposed, to improve the detection and analysis of anomalies within wind turbine operations. By integrating a model of a wind turbine with real-time data from an operational turbine, this framework allows a comparison between the expected and actual operational behaviour. To ensure that the model behaves correctly, the parameters of the model are optimised based on the measurement data and by using intelligent optimisation algorithms to match the behaviour of the real wind turbine. For the optimisation process surrogate optimisation, particle swarm optimisation and a genetic optimisation algorithm are used, and the results are compared to each other.