Acoustical damage detection of wind turbine yaw system using Bayesian network. (November 2020)

• Record Type: Journal Article
• Authors: Chen, Bin; Xie, Lei; Li, Yongzhan; Gao, Baocheng
• Is Part Of: Renewable energy. Volume 160(2020)
• Page Start: 1364
• Page End: 1372
• Publication Date: 2020-11
• Subjects: Wind turbine yaw system -- Damage detection -- Acoustic signal -- Bayesian network; Renewable energy sources -- Periodicals; Power resources -- Periodicals; Énergies renouvelables -- Périodiques; Ressources énergétiques -- Périodiques; 333.794
• DOI: 10.1016/j.renene.2020.07.062
• Languages: English
• ISSNs: 0960-1481

Abstract

Abstract: Yaw system plays a significant role in increasing wind power production and protecting the wind turbine. However, the working yaw system suffers from complex alternating stresses and could result in failure and significant economic losses. This paper develops an acoustical damage detection method of the yaw system based on Bayesian network (BN). In the method, the sound pressure level (SPL) features are first extracted from the measuring acoustic signal to characterize the state of yaw system. Subsequently, a data discretization method based on self-organizing map and information gain rate is proposed to convert continuous SPL features into a finite set of intervals with respect to attribute values. Besides, a three-layer BN diagnostic model combined with the structure learning strategy based on Bayesian information criterion is designed for damage detection of the yaw system. Finally, experiments are conducted in practical wind farm to validate the feasibility and efficiency of the proposed method. Highlights: An acoustical damage detection method of the wind turbine yaw system is proposed. A data discretization method based on information gain rate is presented to process continuous features. A three-layer BN model with hybrid strategies is designed for detecting yaw system. Experimental data from practical wind farm is employed to evaluate efficiency of proposed method.