论文部分内容阅读
导弹自动驾驶仪在振动测试过程中存在信号基线漂移且污染严重的问题,而传统的时频处理方法难以达到消噪要求。因此本文基于广义形态学基本原理提出了一种用于解决振动信号基线漂移的滤波方法。该滤波方法有三级结构组成,前两级结构均是基于形态学基本原理,第三级进行相消与平滑处理,通过相互级联,可以有效抑制基线漂移。此外,通过引入粒子群优化算法(Particle Swarm Optimization,PSO)使得该滤波方法更具适应性。最后的实验中利用本文方法和对比方法对自动驾驶仪实测振动信号与标准ECG(Electrocardiogram)信号进行了处理,结果表明:本文提出的滤波方法在抑制基线漂移方面要优于小波阈值去噪和传统的形态学去噪。
The missile autopilot has the problem of baseline drift and serious pollution in the vibration test process, while the traditional time-frequency processing method is difficult to meet the noise elimination requirements. Therefore, this paper proposes a filtering method based on the principle of generalized morphology to solve the baseline drift of vibration signals. The filtering method has three levels of structure, the first two levels of structure are based on the basic principles of morphology, the third level of descent and smoothing, cascade by each other, can effectively suppress the baseline drift. In addition, the particle swarm optimization (PSO) is introduced to make the filtering method more adaptive. In the last experiment, the measured autopilot vibration signals and standard ECG (Electrocardiogram) signals were processed by the method and the comparative method. The results show that the proposed filtering method is superior to wavelet threshold denoising and traditional Morphological denoising.