College of Information Science and Engineering, Northeastern University, Shenyang 110819, China.
Sensors (Basel). 2022 Sep 1;22(17):6602. doi: 10.3390/s22176602.
Elevator car vibration signals are important information to monitor and diagnose the operating status of elevators, but during the process of vibration signals acquisition, vibration signals are always inevitably disturbed by noise, which affects further research. Therefore, aiming at the vibration signal with noise, we propose a new vibration signal denoising method on the basis of complementary ensemble empirical mode decomposition (CEEMD) and bilateral filtering. Firstly, the collected original vibration signals are decomposed by the CEEMD into several inherent mode functions. Then, the false components are removed by determining the correlation coefficients of mode components, and then the noise-dominate components are denoised by bilateral filtering. Finally, the processed inherent mode functions are reconstructed to require the denoised signal. We test the method through simulation and practical application. The results indicate that the proposed method can efficaciously reduce the noise in the vibration signal of an elevator car.
电梯轿厢振动信号是监测和诊断电梯运行状态的重要信息，但在振动信号采集过程中，振动信号总是不可避免地受到噪声干扰，这影响了进一步的研究。因此，针对带有噪声的振动信号，我们提出了一种基于互补集合经验模态分解（CEEMD）和双边滤波的新的振动信号去噪方法。首先，通过 CEEMD 将采集到的原始振动信号分解为几个固有模态函数。然后，通过确定模态分量的相关系数来去除虚假分量，然后通过双边滤波对噪声主导分量进行去噪。最后，对处理后的固有模态函数进行重构，以获得去噪后的信号。我们通过仿真和实际应用对该方法进行了测试。结果表明，所提出的方法可以有效地降低电梯轿厢振动信号中的噪声。
