节点文献
神经网络及声谱分析技术在刚性路面脱空识别中的应用
Application of Neural Network and Acoustic Charts Analysis Technology on Rigid Pavement Void Identification
【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。
【Abstract】 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.
【关键词】 道路工程;
刚性路面;
神经网络;
声谱分析;
脱空识别;
【Key words】 road engineering; rigid pavement; nerve network; spectrum analysis; void identification;
【Key words】 road engineering; rigid pavement; nerve network; spectrum analysis; void identification;
【基金】 国家自然科学基金资助项目(50078023)。
- 【文献出处】 森林工程 ,Forest Engineering , 编辑部邮箱 ,2005年01期
- 【分类号】U416.222
- 【被引频次】2
- 【下载频次】94