摘要
通过数据采集得到三种不同类型车辆的地震动信号,采用小波消噪和特征提取,得到样本数据对神经网络进行训练,训练完成的神经网络就能实现车辆类型的识别。试验结果表明,BP神经网络对车辆目标具有较高的识别率,证明对地震动信号的特征提取方法是正确的,人工神经网络是有效的目标识别方法。
BP neural network is an important method in target recognition.Wavelet transformation is widely used in signal de-noising and feature extraction.The ground vibration signals of three different vehicles can be got by data acquisition.The wavelet de-noising and feature extraction are adopted to achieve the sample data to train the neural network.The successfully trained neural network can recognize the types of different vehicles.The experimental result shows that BP neural network has high recognition rate for vehicles,the feature extraction method of ground vibration signals is correct,and artificial neural network is an effective method for target recognition.
出处
《现代电子技术》
2012年第10期100-102,共3页
Modern Electronics Technique
关键词
BP神经网络
小波变换
目标识别
数据融合
BP neural network
wavelet transformation
target recognition
data fusion
