论文部分内容阅读
针对车辆噪声识别问题，在对实测的４种车辆噪声信号进行预处理和ＡＲ谱估计的基础上，提取了谱重心、谱面积和卓越频率成分等几个特征量。实测的车辆噪声数据检验表明，基于ＡＲ谱结构提取的这几个特征具有很好的稳定性，且对车辆目标具有良好的识别效果，这对最终完成战场声目标模糊模式识别具有重要的参考价值。
Aiming at the problems of vehicle noise identification, several characteristic parameters of spectrum center of gravity, spectrum area and excellent frequency are extracted based on the preprocessing and AR spectrum estimation of the measured four kinds of vehicle noise signals. The measured vehicle noise data show that these features extracted based on the AR spectral structure have good stability and good recognition of the vehicle targets, which is of great reference value for the final completion of the fuzzy target pattern recognition of battlefield .