摘要
利用AR参数模型提取采集到的车辆行驶时产生的车外噪声信号的特征,利用假设检验进行特征选择,并设计了BP神经网络进行分类识别.对道路现场采集到的2种车型共计110个信号进行分析,实验结果表明,通过AR参数模型提取车辆车外噪声特征实现车型自动分类是可行的,其分类的正确率达80%以上.
This paper extracted the futures of collected radiated noise of running vehicles by AR model firstly, then opted features by hypothesis test, and carried out classification by BP-NN. This papes analyzed 110 signals that are colleted on road in spot, the results indicated that it is feasible that AVC （Automatic Vehicle Classification） can be realized by features of radiated noise that are extracted by AR mode/, it＇s accuracy is more than 80 percent.
出处
《武汉理工大学学报（交通科学与工程版）》
2008年第6期1056-1058,1062,共4页
Journal of Wuhan University of Technology（Transportation Science & Engineering）
基金
国家自然科学基金项目资助(批准号:50478007)
