多抽样率原理在舰船振动信号分析中的应用与研究

Multi-rate spectrum octave analysis of ship vibration

船舶航行过程中产生的振动会影响船上人员的舒适性、降低船体结构耐久性并影响舰船的隐身性能。船体振动评估对于船舶的防振隔振设计、确定振动量级并判断是否满足衡准要求具有重要意义。随着技术的进步和使用要求的提高,对船体振动评价的客观性和全面性也越来越高。倍频程分析是振动分析中常用的频率分析方法,早期的方法是带通滤波器法,需要采用大量的滤波器以提高频率分辨率,导致计算量大运算时间长。为此,提出了基于多抽样率原理的倍频程频域方法。论文对该频域方法进行了详细介绍,对方法的谱分辨率、计算量和存储空间进行了分析对比,基于FPGA搭建了船舶振动采集分析系统,在FPGA中对算法进行了实现,最后对实测船舶振动信号应用多抽样率倍原理进行了倍频程分析。

Vibration is one of the main factors affecting the habitability of ships, degrading the durability of the ships＇ structure and stealth performance as well. Estimating the vibration of ships is important for both vibration isolated designing and evaluating whether the magnitude of vibration meets the criterion. The requirements of objectivity and comprehensiveness for ship vibration evaluation become more and more demanding. Octave frequency spectrum analysis method is a commonly used in vibration analysis. The bandpass filter bank method is commonly used to implement octave analysis, but that incurs heavy computation when large amounts of filtering calculation are needed in order to improve the frequency resolution. Therefore, a multi-rate octave analysis method is proposed and introduced in detail. Resolution, computational complexity and memory space of the two methods are compared. The algorithm is implemented in FPGA based ship vibration data acquisition system, and the collected vibration data is analyzed via multi-rate spectrum octave method.