Affiliations: [a] SPIC Jiangsu New Energy Co., Ltd., Yancheng, Jiangsu, China | [b] Shanghai Power Equipment Research Institute Co., Ltd., Shanghai, China
Correspondence:
[*]
Corresponding author: Chao Fang, Shanghai Power Equipment Research Institute Co., Ltd., Shanghai 200240, China. E-mail: fangchao_vip@outlook.com.
Abstract: Although the global wind energy industry has made considerable progress in recent years, wind turbines suffer from frequent failures since the systems are complicated and the working conditions are far from being satisfactory. For the wind turbines to function well, it is imperative to study the overall status of the wind turbine unit, evaluate the performance of the wind farm, apply intelligent operation and maintenance technology, and improve operation and maintenance strategies on an on-going basis, all of which are based on the operation data of the unit. This paper focuses on the evaluation and extension of the lifetime of wind turbines. Based on relevant knowledge and theories from previous studies, an evaluation method based on big data was designed to do the evaluation, and the results of which were verified by real cases. With the duration of catastrophic failures taken into account, the proposed life prediction algorithm was proved to be effective. If the bearing runs for 34 days, the actual remaining life of wind turbines is 0.2 days. The number predicted for LRM is 0.8 days and that predicted for ILRM is 0.31 days. Compared with LRM, the prediction for ILRM is much more accurate.
Keywords: Big data, wind power generation, lifetime evaluation, optimized operation
