摘要
对风力发电机组滚动轴承故障进行准确诊断,可以对故障进行及时维修,提高风力发电机组滚动轴承的使用寿命。进行故障诊断时,需要对滚动轴承故障类型进行分类后,并与数据库中已知的故障类型进行比对,而传统方法使用经验模式对故障信号进行分解,提取故障信号的特征向量完成诊断,但不能对故障类型进行分类和比对,故障诊断时间较长,精度低、效果差。提出一种新的时频维数的风力发电机组滚动轴承故障诊断方法。依据多重分形的定义及分形几何学相关原理,依据滚动轴承时频维数确定轴承故障类型,并将同数据库中的数值进行对比,实现滚动轴承故障的检测。仿真结果表明,改进方法可以对风力发电机组滚动轴承是否发生故障进行准确的判断,在风力发电机组滚动轴承故障发生初始时间就可以进行诊断和报警,同时还能对滚动轴承故障的类型进行准确判断,实用性强。
It can repair the fault timely and rolling bearing service life of wind generating set to diagnose its fault accurately. The fault diagnosis needs to classify the fault types and compare the known fault types in database. However, traditional method decomposes the fault signal using experience mode and extracts the feature vector of fault signal to complete the diagnosis. It cannot clasSify and compare the fault types and leads to long diagnosis time and poor accuracy and effect. In this paper, we propose a fault diagnosis method for rolling bearing of wind generating set based on new time -frequency dimension. It confirms the fault types according to the time -frequency dimension of rolling bearing based on multi - fractal definition and correlation theory of fractal geometry. It also compares the value in the same database to achieve the fault diagnosis of rolling bearing. The simulation results show that the modified method can judge whether the rolling bearing breaks down accurately and can also judge the rolling bearing fault types accurately in the meantime. It has good practicability.
出处
《计算机仿真》
CSCD
北大核心
2016年第12期105-108,共4页
Computer Simulation
关键词
时频维数
滚动轴承
故障诊断
Time - frequency dimension
rolling bearing
Fault diagnosis
