基于遗传算法的信号识别技术  

Signal Recognition Based on Genetic Algorithm

作　　者：陶继忠[1] 殷国富[1] 汪法根[2] 

机构地区：[1]四川大学制造科学与工程学院,成都610065 [2]中国工程物理研究院机械制造工艺研究所,绵阳621900

出　　处：《振动．测试与诊断》2004年第3期176-178,共3页Journal of Vibration,Measurement & Diagnosis

基　　金：国家自然科学基金资助项目 (编号 :5 0 2 75 1 0 0 )

摘　　要：将遗传算法应用于识别气浮式动平衡测量装置准静态信号的参数 ,提出了二次磨光法的数据预处理方法。首先采用二次磨光法消除坏点并进行快速傅利叶变换以确定信号的频率成分 ,并设定各频率成分信号的幅值和初相位的取值范围 ,然后以实测信号与识别结果的差值定义适应度 ,最后经过基本遗传操作和迭代得出最优解。仿真结果和试验实例表明 ,该方法精度很高 ,幅值识别误差不超过 1 %,初相位误差小于 2°,满足了实际工程需求。A new application of GA to the recognition of guasi static signal parameters of an air suspending dynamic balance measuring device is presented.A data preprocessing method based on the quadric finishing function is suggested,and the effect of GA parameters on the results is analyzed.Firstly,the FFT is used to analyze the frequency components of the signals,including their amplitudes and phases,and the finite intervals over which the values of the amplitude and initial phase angle of each frequency component are to be taken are set up.Secondly,an adaptive degree function is defined by the difference between measured and recognized signals.Finally,Optimal Solutions are obtained through basic genetic operations and iteration.Simulation and test results show that the method presented in the paper is quite accurate with maximum error of 1% in amplitudes,and 2° in phases angles,which satisfies engineering demands.

关 键 词：遗传算法 信号识别 格雷码 二次磨光法 相位误差 

分 类 号：TN911.7[电子电信—通信与信息系统]