论文部分内容阅读
针对付氏分析对时变信号不敏感和缺乏空间局部性的弱点，提出了一种基于正交小波变换的原始特征参数的构造方法，在此基础上通过遗传算法提取蕴涵于小波系数序列的故障特征，获得较理想的特征参数。实验证明，本文提出的这种方法用于特征提取与模式识别是有效的。
Aiming at the weakness of Fu’s analysis on time-varying signals and lacking of spatial locality, a method of constructing original feature parameters based on orthogonal wavelet transform is proposed. Based on this, a genetic algorithm is proposed to extract the fault contained in the wavelet coefficient sequence Characteristics, access to the more ideal characteristics of parameters. Experiments show that this method proposed in this paper is effective for feature extraction and pattern recognition.