改进加权融合算法与双谱技术在风电轴承故障诊断中的应用

Application of Improved Weighted Fusion Algorithm and Bispectrum Technique in Wind Turbine Bearing Fault Diagnosis

针对风电回转支承转速低、不稳定、受载复杂等问题,并且其故障振动信号为低频信号,往往呈现非高斯、非线性特点。提出了基于改进加权融合算法与双谱技术相结合的数据融合诊断方案。首先提出了改进的加权融合算法对多传感器数据进行融合;然后采用双谱技术处理非高斯、非线性信号,提取出隐藏在融合信号中的非高斯特征。最终通过实验表明,运用该诊断方案提取出的特征频率与理论计算的故障特征频率相吻合,充分显示了其应用在回转支承故障诊断系统中的可行性。

Wind turbine bearing with low speed is unstable and complex,and its fault vibration signal has characteristics such as low-frequency,non-Gaussian,non-linear and so on.In view of these problems,it puts forward a project based on improved weighted fusion algorithm and bispectrum technology.First of all,it proposes the improved weighted fusion algorithm to fuse the multi-sensor data.Then,bispectrum technique is used to extract non Gauss features hidden in the fused signal.Finally,the experimental results show that the characteristic frequency extracted by the diagnosis scheme is consistent with the calculated characteristic frequency of the fault,which fully shows the feasibility of the application in the fault diagnosis system.