摘要
通过简单的敲击法获取刚性路面声频信号,利用MATLA B软件中的信号分析与处理工具箱进行音频信号的声谱分析并定义音频变量和音频峰值等状态参数,应用多层前馈神经网络系统和Dempster-shaft理论,实现刚性路面脱空的判定。
The acoustic signals are captured by means of simple knock on rigid pavement. In the signal analysis and processing toolbox of MATLAB software, the spectrum analysis of acoustic signals are carried through, and the audio frequency variable and audio frequency peak value are defined. Through the application of Multilayer feed-forward neural network system and Dempster-shaft theory, the void identification of rigid pavement is implemented.
出处
《森林工程》
2005年第1期20-23,共4页
Forest Engineering
基金
国家自然科学基金资助项目(50078023)。
关键词
道路工程
刚性路面
神经网络
声谱分析
脱空识别
road engineering
rigid pavement
nerve network
spectrum analysis
void identification
