摘要
针对大型机械设备运行环境恶劣故障特征难以提取的问题,提出一种自适应二阶双稳态随机共振方法。首先系统输出信号的信噪比作为蚁群算法的自适应度函数,然后采用蚁群算法优化二阶随机共振系统的参数和阻尼因子,再利用优化得到的最佳参数设置二阶随机共振系统,最后实现微弱故障特征的增强与提取。数值仿真分析表明:该方法可以有效地提取淹没在强噪声背景下的微弱正弦信号;而且深沟球轴承滚动体故障实验结果证明提出的方法能有效增强与提取滚动体故障特征频率。仿真与实验对比结果表明:提出的方法优于传统随机共振方法,归功于该方法不仅能够利用蚁群算法并行选择和优化随机共振系统参数,而且克服传统随机共振方法对高通滤波器的依赖。
An adaptive second-order bistable SR method is proposed to extract bearing fault characteristics in heavy background noise. First, the signal to noise ratio （SNR） of output signal of second-order stochastic resonance system is set as the objection function of colony algorithms. Second, the colony algorithms are employed to select and optimize the system parameters and damping factor. Finally, the optimal parameter pair is used to set the second-order stochastic resonance system to enhance and extract the bearing fault characteristics. Simulation data indicate that the proposed method can effectively extract weak characteristics in heavy background noise. Rolling element bearing case with an incipient roller fault demonstrates that the proposed method possesses strong enhancement capability and is superior to the existing first-order SR methods. The reason is that the proposed method not only selects system parameters adaptively by using colony algorithms, but also is independent on the help of highpass filters.
出处
《中国测试》
北大核心
2017年第6期31-36,共6页
China Measurement & Test
关键词
二阶双稳态随机共振
蚁群算法
微弱特征检测
故障诊断
second-order bistable stochastic resonance
colony algorithms
weak characteristic detection
fault diagnosis
作者简介
罗毅（1980-），男（彝），贵州贵阳市人，讲师，硕士，研究方向为电子信息、信号处理、自动控制。
