摘要
以风电机组齿轮箱为研究对象,提出似然学习机学习算法进行风电机组齿轮箱状态监测技术研究。首先从风电机组齿轮箱振动信号中提取峭度为特征量,然后利用似然学习机对特征量进行学习生成高斯模型,最后利用高斯模型对齿轮箱进行状态评估。利用风电机组传动系统实验平台和美国凯斯西储大学实验进行验证,验证了该文提出方法的有效性。
Taking wind turbine gearbox as the research object,the likelihood learning machine algorithm is proposed to study the status monitoring technology of wind turbine gearbox.Firstly,the kurtosis is extracted from the vibration signal of the wind turbine gearbox as feature.Then the likelihood learning machine is used to learn the feature to generate the Gaussian model.Finally,the Gaussian model is used to evaluate the state of the gearbox.The effectiveness of the proposed method is verified by using the wind turbine driving system test platform and the experiment of Case Western Reserve University.
作者
李东东
华伟
郑小霞
赵耀
王浩
Li Dongdong;Hua Wei;Zheng Xiaoxia;Zhao Yao;Wang Hao(College of Electrical Engineering,Shanghai University of Electric of Power,Shanghai 200090,China;Shanghai Higher Institution Engineering Research Center of High Efficiency Electricity Application,Shanghai 200090,China;College of Automation Engineering,Shanghai University of Electric of Power,Shanghai 200090,China;State Grid of Jiangxi Electric Power Overhaul Branch,Nanchang 330000,China)
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2021年第4期374-379,共6页
Acta Energiae Solaris Sinica
基金
国家自然科学基金(51507098)
上海市科学技术委员会(17DZ2293500
17020500800)
上海市“曙光计划”(15SG50)。
关键词
齿轮
风电机组
状态监测
似然学习机
峭度
高斯模型
gear
wind turbine
status monitoring
likelihood learning machine
kurtosis
Gaussian model