Predictive maintenance of wind turbines based on digital twin technology

Abstract

Predictive maintenance of wind turbines plays a crucial part in directing power grid dispatching and maintaining power grid security. In this paper, a way of ultra-short term wind power prediction relied on digital twin technology is proposed, which realizes actual time and accurate wind power prediction by building a digital model. Firstly, BP neural network (back propagation neutral network) was used to predict the wind power, and the initial predicted value of wind power was obtained. Then the meteorological data was substituted into the historical meteorological database to find similar meteorological conditions data, and the BP neural network predicted value was weighted to get the final digital twin predicted value. The simulation results indicate that this method can effectively enhance prediction accuracy of wind power.

Keywords

• Predictive maintenance
• BP neural network
• Digital twin
• Prediction accuracy