改进相关性消噪算法在铣削力信号消噪中的应用  

Milling Force Signal Denoisied by Improved Correlation Algorithm

作　　者：刘木森 王杰[1] 刘欢[1] 蔡明浩 Liu Musen;Wang Jie;Liu Huan;Cai Minghao

机构地区：[1]四川大学

出　　处：《工具技术》2018年第11期83-89,共7页Tool Engineering

基　　金：四川省重点科技计划(2017GZ0092)

摘　　要：铣削加工状态的判定对保证铣削加工产品质量具有重要意义。铣削力信号中包含有丰富的铣削加工状态信息,但其中混杂的噪声影响力信号的准确性。传统的小波变换尺度间相关性去噪算法运算量大、效率低、准确性差,不适合实时铣削力信号消噪。通过引入系数比较权重、噪声能量权重及从大尺度到小尺度提取边缘信息的方法,对传统算法进行改造,提高了铣削力信号消噪效率及准确性。将改进前后的算法和阈值去噪算法分别用于对低信噪比和高信噪比信号进行消噪处理,结果表明,改进后的算法消噪精度更高,能够保存更多的信号能量。The status of a milling process is significant to the milling quality. Abundant information on a milling state is included in the milling force signal which inevitably contains noises decreasing the signal accuracy. Based on the correla- tion of different scales＇ wavelet coefficients,the traditional denoising algorithm is low in execution efficiency and accuracy. Thus, a novel algorithm using the coefficient comparison and noise energy weights, and the sahation point extraction from a large scale to a small scale are proposed. Compared with the traditional and the threshold denoising algorithms, the accuracy of the modified algorithm is improved. Through simulation, the superiority of the improved algorithm is testified in execution efficiency and preserved signal energy and details.

关 键 词：消噪 小波变换 相关性算法 铣削力 

分 类 号：TG806[金属学及工艺—公差测量技术] TG501.3