论文部分内容阅读
交叉证认技术改进了基于经验模态分解(EMD)的激光雷达回波信号降噪方法。该方法在对激光雷达回波信号噪声特性和经典降噪方法缺陷进行研究的基础上,利用交叉证认技术自适应地识别雷达回波信号中的信号层和噪声层,再通过经验模态分解算法分离噪声和重构信号。通过仿真数据和实测雷达信号对比分析,该方法能够自适应地选择本征模函数中的信号层数,不但有效地滤除了各种随机噪声,而且保留了信号的有效信息特征,减少了信号损失,进而提高了后续数据处理的准确度。
Cross-validation technique improves the lidar echo signal denoising method based on Empirical Mode Decomposition (EMD). Based on the study of the noise characteristics of laser radar echo signal and the defect of classical noise reduction method, this method uses cross-validation technique to adaptively identify the signal and noise layers of radar echo signals and then decomposes them by empirical mode decomposition The algorithm separates noise and reconstructs the signal. By comparing the simulation data with the measured radar signal, this method can adaptively select the number of signal layers in the eigenmode function, not only effectively filtering out various random noise, but also preserves the effective information features of the signal and reduces the signal loss , Thus improving the accuracy of subsequent data processing.