摘要
为了实现国家“2035碳达峰”的战略发展目标,风能作为清洁可再生能源之一受到广泛关注。但大型风电装备结构复杂,工况恶劣,且维护不便,其内部零件随时间的推移可能会发生故障,风场急需配备完善的预测与健康管理系统。现有风场通常分别安装SCADA系统与CMS系统来监测风机的整体运行状态和传动链振动异常,2套系统在数据采样频率与存储策略中存在差异,如“数据孤岛”现象,无法联合诊断分析。因此,提出一种缓速变参数融合的风电剩余寿命预测方法,该方法能将不同类型数据的特点综合到一起,使所得信息更优质,提高了剩余使用寿命(RUL)的预测精度。
In order to achieve the national“2035 carbon peak”strategic development goals,wind energy as one of the clean renewable energy has been widely concerned.However,the structure of large-scale wind power equipment is complex,the working conditions are harsh,and its internal parts will inevitably fail over time.Therefore,the wind farm urgently needs to be equipped with a perfect forecasting and health management system.In existing wind farms,SCADA system and CMS system are usually installed respectively to monitor the overall operating status of the fan and abnormal vibration of the transmission chain.The data sampling frequency and storage strategy of the two systems are very different,and there is a phenomenon of“data island”,which cannot be jointly diagnosed and analyzed.Therefore,a method for wind power residual life prediction based on slow/fast parameter fusion is proposed.This method can integrate the characteristics of different types of data together,so as to obtain better information and improve the prediction accuracy of remaining useful life(RUL).
作者
达文刚
文成龙
孟莉
孙若斌
陈雪峰
DA Wengang;WEN Chenglong;MENG Li;SUN Ruobin;CHEN Xuefeng(Xinjiang Huadian Weihuliang New Energy Co.,Ltd.,Changji 830039,Xinjiang,China;School of Future Technology,Xi’an Jiaotong University,Xi’an 710049,Shaanxi,China;School of Mechanical Engineering,Xi’an Jiaotong University,Xi’an 710049,Shaanxi,China)
出处
《中国工程机械学报》
北大核心
2025年第5期801-805,811,共6页
Chinese Journal of Construction Machinery
基金
国家自然科学基金青年科学基金资助项目(52105118)。
关键词
剩余寿命预测
缓/速变参数融合
贝叶斯更新
residual life prediction
slow/fast parameter fusion
bayesian updating
