CAJ | 학술논문

为解决对非平稳振动信号直接进行循环维纳滤波噪声去除效果较差的问题，将离散余弦变换应用于循环维纳滤波中，提出振动信号离散余弦变换域循环维纳滤波算法。利用离散余弦变换能量集中的特性和循环维纳滤波线性相位滤波特性，采用分段处理方法使二者结合，以便更充分发挥维纳滤波的性能，实现更好的滤波效果，并降低处理阶次。结合工业现场实测汽轮机振动信号，对算法进行分析与仿真，仿真结果表明：该算法拥有较好的噪声滤除效果，且能保证滤波后不产生明显相位失真，适于振动信号去噪处理，优于单独进行循环维纳滤波或DC T滤波。
위해결대비평은진동신호직접진행순배유납려파조성거제효과교차적문제，장리산여현변환응용우순배유납려파중，제출진동신호리산여현변환역순배유납려파산법。이용리산여현변환능량집중적특성화순배유납려파선성상위려파특성，채용분단처리방법사이자결합，이편경충분발휘유납려파적성능，실현경호적려파효과，병강저처리계차。결합공업현장실측기륜궤진동신호，대산법진행분석여방진，방진결과표명：해산법옹유교호적조성려제효과，차능보증려파후불산생명현상위실진，괄우진동신호거조처리，우우단독진행순배유납려파혹DC T려파。
To solve the problem that the non-stationary vibration signals’ noise can not be directly removed by using the circulation Wiener filter ,the paper proposes the discrete cosine transform domain’ s cycle Wiener filtering algorithm for vibration signals applying the discrete cosine transform to the circulation Wiener filter .By using the characteristics of the energy concentration of discrete cosine transform and the cyclic linear phase filter of wiener filtering ,and combining both characteristics by the segmented processing method ,a better filtering effect and a lower order processing are achieved .Based on the industrial field measurement of turbine vibration signals ,the algorithm is analyzed and simulated . The simulation result shows that the algorithm has good noise filtering effect ,and can guarantee no significant phase distortion after filtering . This algorithm is suitable for removing noise of vibration signals and better than the circular Wiener filter or discrete consine T ransform filter .