基于插值重采样的信号FFT分析方法研究

针对实际信号做频谱分析时通常不能满足经典FFT算法的待分析信号长度应为2的整数幂的约束条件,通常采用的增加信号采样时间或对信号补零的方法,会造成采样负荷加重或频率泄露。针对与此,提出了基于信号插值重采样的改进FFT分析算法,并与常用FFT算法进行了对比。经仿真结果验证,改进的算法能较好地在不增加信号采样负荷的前提下,精确捕获信号真实频率特性。并将改进的FFT算法应用于实际水电机组振动信号分析,频谱分析与现场真实情况相符,证明基于插值重采样的FFT算法的有效实用性。

Research on Fast Fourier Transform Algorithm Based on Interpolation Re-sampling Method

Regarding that the actual signal while its frequency spectrum is analyzed usually doesn’t meet the demand of radix-2 Fast Fourier Transform(FFT) which is only worked for the sequence length that is an integral power of 2,the methods of prolonging the sampling time or padding zero to the signal so that it fits for the algorithm are adopted,but these methods aggravate sampling load and cause omission of frequency at the same time.An improved FFT algorithm based on the linear interpolation method is given and compared with the common one herein.The simulation results show that the improved algorithm can precisely get the frequency characteristics of the signal without adding workload.The application of improved algorithm on the actual analysis of hydroelectric set vibration signal also shows that the results are consistent with the real situation.