基于傅立叶变换和奇异值分解的路面脱空判定

Pavement Void Identification Based on Fourier Transform and Singular Value Decomposition

声振法检测水泥混凝土路面脱空所采集的是声振信号,可用频谱图分析,采用奇异值分解可有效地提取特征矢量,并将特征矢量送入神经网络进行模式识别。分析结果表明,能在车辆噪声干扰下有效地识别水泥混凝土路面板下地基的脱空情况,为检测路面板下地基脱空的方法提供了理论依据。

Acoustic vibration method detects cement concrete pavement void by collecting acoustic signals for spectrum analysis.Characteristic vector can be effectively extracted using singular value decomposition and then sent to neural network for pattern recognition.The result shows that it can effectively identify foundation void of cement concrete pavement,which provides theoretical basis for detecting foundation void.