【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,再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。本文对道路现场采集到的两种车型共计74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。更多还原

【Abstract】 The features of collected radiated noise of driving vehicles were extracted by means of the AR parametric model.They were selected by the hypothesis test and classified by the designed BP neural network.74 signals of two types of vehicle collected on road in situ were analyzed.The results show that it is feasible to realize the automatic vehicle classification by the features of radiated noise extracted by the AR model,and the correctness rate of the classification is higher than 80%.更多还原