激光信道的随机性较强,对环境影响因素变化较敏感,导致激光雷达回波传输中存在一定的噪声,形成了包含不平稳数据的未知环境。为此,提出一种激光雷达非平稳数据隔离方法,及时隔离未知环境中的相关数据,保证输出传输质量。将干扰下的激光雷达回波信号分割为多个不同的子集,降维和去噪处理激光雷达回波数据。以互信息获取和与最 相关的变量为脉冲能量,获取回波信号特征变量,以此为依据利用支持向量数据描述(SVDD)方法检测非平稳数据。利用颜色-时序标记对象(CTMO)模型在全部回波数据与非平稳数据之间建立影响关系,利用该关系实现激光雷达非平稳数据隔离。实验结果表明,所提方法可以准确检测激光雷达非平稳扫描点、离群值和局部异常值三种数据,最高计算时间为25s;颗粒物激光雷达接收机非平稳数据噪声基底与实际频谱变化基本一致,保留了传输数据的细节特征;数据传输过程中的最高时间负载增量为4.8%,可有效隔离非平稳数据。

The randomness of the laser channel is strong, and it is sensitive to the change of environmental factors, resulting in certain noise in the transmission of the laser radar echo, forming an unknown environment with unstable data. For this reason, a method of laser radar nonstationary data isolation is proposed to isolate the relevant data in the unknown environment in time to ensure the output transmission quality. The lidar echo signal under jamming is divided into several different subsets, and the dimension reduction and denoising are used to process the lidar echo data. The characteristic variable of the echo signal is obtained by taking the pulse energy as the most relevant variable of mutual information acquisition and detection, and the nonstationary data is detected by support vector data description (SVDD) method. The color-time marks object (CTMO) model is used to establish the influence relationship between all echo data and nonstationary data, and the nonstationary data isolation of lidar is realized by using this relationship. The experimental results show that the proposed method can accurately detect the nonstationary scanning points, outliers and local outliers of lidar, and the maximum calculation time is 25 s. The noise base of the nonstationary data of the particle laser radar receiver is basically consistent with the actual spectrum change, and the detailed characteristics of the transmitted data are retained. The maximum time load increment during data transmission is 4. 8??, which can effectively isolate nonstationary data.

TN958.98

高功率激光器实验室开放基金资助项目(2018KIE93); 国家自然科学基金资助项目(61301136);ISN 国家重点实验室开放课题资助项目(ISN21-07)