基于实测冲击响应字典稀疏表示的齿轮系统侧隙故障特征提取  

Backlash fault feature extraction of spur gear systems based on sparse representation dictionary of measured impulse responses

作　　者：高洪波[1,2] 刘杰[1] 李允公[1] GAO Hong-bo LIU Jie LI Yun-gong(School of Mechanical Engineering ＆ Automation，Northeastern University，Shenyang 110819 Department of Electromechanical Engineering, Liaoning Provincial College of Communications，Shenyang 110122)

机构地区：[1]东北大学机械工程与自动化学院,沈阳110819 [2]辽宁省交通高等专科学校机电工程系,沈阳110122

出　　处：《振动与冲击》2017年第11期86-91,共6页Journal of Vibration and Shock

基　　金：国家自然科学基金(51275080)

摘　　要：为了从齿轮系统振动信号中提取由于齿侧间隙故障产生的瞬态冲击成分,提出了基于稀疏表示的故障特征提取方法,将特征提取转化为稀疏字典的构造和稀疏表示系数的求解。根据侧隙故障瞬态冲击的响应特性,提出采用与其更相似的系统实测冲击响应构造稀疏字典,继而利用正交匹配追踪算法优化求解稀疏表示向量,获得表征瞬态冲击发生时刻的稀疏系数,并重构故障冲击序列。通过冲击成分的包络解调分析,进一步获得瞬态冲击的调制频率。仿真和实验表明,提出的方法能够有效提取齿轮系统齿侧间隙故障特征。To separate the transient impact responses caused by backlash fault from the vibration signal of gear system，a feature extraction method based on sparse representation is presented, in which feature extraction is converted to the construction of sparse dictionary and the solution of sparse coefficients. According to its more similar characteristic to fault impact response, using the measured impulse response as an atom to construct sparse dictionary is proposed. Then the orthogonal matching pursuit algorithm is employed for the optimal solution of sparse vectors and the reconstruction of impact response sequence, and the occurrence time of transient impact represented by sparse coefficients are obtained. Furthermore, the modulation frequency is demodulated by envelope analysis of the impact response composition. Simulation and experiments show that the method proposed is effective for backlash fault feature extraction of gear system.

关 键 词：故障特征提取 稀疏表示 实测冲击响应 齿侧间隙 

分 类 号：TH17[机械工程—机械制造及自动化]