摘要
为提高对指挥信息系统有线远距离传输接口测试的自动化水平,提出一种基于信号特征提取的信号类型自动识别方法;在对各类接口信号时域和频域特性分析的基础上,采用限幅滤波法实现对信号峰峰值的准确判断,通过基于Burg算法AR模型的谱估计方法实现信号频谱准确估计;依据提取到的信号峰峰值和频带宽度,采用双阀值判决方法对接口类别进行自动识别;实验结果表明,基于信号时频域特征差异的自动识别方法,能够有效区分3种类型15个接口,识别准确度达到87%以上。
To improve the long--distance transmission of command information systems wired interface test automation level, proposed the signal automatic identification method based on the type of signal feature extraction. On the basis of various types of interface signals in time domain and frequency domain characteristics analysis, the use of limiting filtering method to achieve peak to peak signal accurate judgment, for signal spectrum by Burg spectral estimation method based on AR model algorithm for accurate estimates. Based on the extracted signal peak to peak and bandwidth, dual threshold decision method for automatic recognition interface categories. Experimental results show that the automatic recognition method characterized in the frequency domain based on the difference signal, can effectively distinguish three types of 15 interfaces, recognition accuracy reaches 87%.
出处
《计算机测量与控制》
2016年第1期255-258,共4页
Computer Measurement &Control
基金
总装科研项目
关键词
接口信号识别
特征提取
限幅滤波法
谱估计
interface signals recognition
feature extraction
limiting filtering
spectral estimation
