摘要
提出基于小波信号特征提取BP神经网络算法目标预测识别的方法;利用振动传感器采集地面移动目标的振动信号,再通过小波变换算法分析出振动信号的特征,通过BP神经网络实现对不同类型移动目标的识别;文章介绍了小波变换理论以及BP神经网络算法,详细阐述了信号特征提取以及利用BP神经网络进行识别和预测;通过实验表明,文章的方法能够快速有效地识别移动物体,人员、小车和大车等移动目标正确识别率达到了94%、84%和88%,并且稳定性好,具有很强的实用性。
This paper proposed that the BP neural network algorithm for recognition of target prediction method based on wavelet signal feature extraction. Using vibration sensors to collect ground moving target of vibration signal, through the Wavelet transform algorithm to analyze the characteristics of the signal by the neural network to achieve the identification of the different types of mobile targets. This article introduces the theory of wavelet transform and BP neural network algorithm, elaborate signal feature extraction using BP neural network to identify and predict. The experiments show that the proposed method can be quickly and efficiently identify moving objects, Personnel, trolley and carts and other mobile targets correct recognition rate reached 94%, 84% and 88%, and good stability, strong practicability.
出处
《计算机测量与控制》
北大核心
2013年第9期2550-2553,共4页
Computer Measurement &Control
基金
四川理工学院科研基金项目资助(2011KY05)
关键词
目标识别
WSNS
小波变换
神经网络
特征提取
Target recognition
WSNs
Wavelet transform
Neural network
Feature extraction
