改进灰狼优化模糊核聚类在风电齿轮箱故障诊断中的应用

Application of Improved Grey Wolf Optimization KFCM Algorithm in Fault Diagnosis of Wind Turbine Gearbox

为准确地辨识已知、未知故障类别,提出一种基于模糊核聚类模型的风电齿轮箱故障诊断新方法。首先,将模型初始聚类中心和核参数作为优化变量,采用改进型灰狼优化算法寻优求解。改进型灰狼优化算法中引入莱维飞行策略和非线性收敛向量,能够提高算法的收敛速度与精度,从而获得最佳分类结果下的各聚类中心和核参数;然后,根据待测样本与各聚类中心之间的核空间样本相似度,先判断样本是否属于已知故障,再诊断故障类别;最后,通过模拟风电齿轮箱的故障实验验证了该方法的有效性。

In order to accurately identify the known and unknown fault types,a new fault diagnosis method for wind turbine gearbox based on the kernel fuzzy c-means clustering(KFCM)is proposed. The initially cluster centers and the kernel parameter of the KFCM model are taken as optimization variables,and an improved grey wolf optimization algorithm is used to find the optimal centers. The introduction of Levy flight strategy and nonlinear coefficient vector in the improved grey wolf optimization algorithm can improve the convergence speed and accuracy of the algorithm,and the clustering centers and kernel parameters can be obtained under the optimal classification results. Then,according to the similarity between the new sample and the centers in the kernel space,firstly whether the sample belongs to a known fault type is determined,and then the fault type is diagnosed. Finally,the effectiveness of the proposed method is verified by experiments simulating different fault types of wind turbine gearbox.