摘要
针对风电机组齿轮箱故障信号受多噪声、多转速影响难以处理的问题,提出一种基于优化变分模态分解(VMD)的改进对称点图(ISDP)和膨胀胶囊网络(DCapsNet)结合的故障诊断方法。首先,提出利用均方根误差和皮尔逊相关系数优化VMD最佳分解数量和惩罚因子的方法,并利用优化后的VMD对故障信号降噪;其次,将去噪后的故障信号转化为多通道多间隔的ISDP;最后,将ISDP输入DCapsNet进行训练。实验结果表明,所提ISDP-DCapsNet方法相比于其他故障诊断方法,具备良好的精确性和有效性。
To address the challenge of processing fault signals from wind turbine gearboxes,which are affected by multiple noise sources and varying rotational speeds,the paper proposes a fault diagnosis method combining improved symmetric dot pattern (ISDP) and dilated capsule network (DCapsNet) based on optimized variational mode decomposition (VMD).Firstly,a method utilizing root mean square error and Pearson correlation coefficient is introduced to optimize the number of decomposition modes and the penalty factor in VMD.The optimized VMD is then applied to denoise fault signals.Subsequently,the denoised fault signals are transformed into multi-channel,multi-interval ISDP.Finally,the ISDP are input into DCapsNet for training.Experimental results demonstrate that the proposed ISDP-DCapsNet method exhibits superior accuracy and effectiveness compared to other fault diagnosis methods.
作者
李俊卿
韩小平
黄涛
张承志
刘若尧
何玉灵
刘雨田
LI Junqing;HAN Xiaoping;HUANG Tao;ZHANG Chengzhi;LIU Ruoyao;HEYuling;LIU Yutian(North China Electric Power University,Baoding 071000,China;State Grid Puyang Power Supply Company,Puyang 457000,China;State Grid Shaanxi Ultra High Voltage Company,Xi’an 710026,China)
出处
《智慧电力》
北大核心
2025年第3期27-34,共8页
Smart Power
基金
国家自然科学基金资助项目(52177042)。
