By Doppler weather radar and satellite cloud data,the Meiyu front rainstorm which occurred in the north of Huaihe River and the north region of Yangtze-Huaihe was analyzed during July 1-2,2006.The characteristics of i...By Doppler weather radar and satellite cloud data,the Meiyu front rainstorm which occurred in the north of Huaihe River and the north region of Yangtze-Huaihe was analyzed during July 1-2,2006.The characteristics of infrared satellite cloud,radar basic reflectivity and Doppler radial velocity were discussed in this rainstorm,such as the echo intensity,the characteristics of velocity field and so on.The reasons for this rainstorm were analyzed by these characteristics.The combination of radar and satellite data in the role of forecasting Meiyu front rainstorm was studied.展开更多
多传感器建图与定位SLAM系统(simultaneous localization and mapping)在室外长距离跨度环境中,由于各传感器信息融合不正确、特征匹配错误,或传感器状态信息不可信,导致建图精度不足,轨迹漂移甚至建图崩溃。对此,提出一种基于因子图优...多传感器建图与定位SLAM系统(simultaneous localization and mapping)在室外长距离跨度环境中,由于各传感器信息融合不正确、特征匹配错误,或传感器状态信息不可信,导致建图精度不足,轨迹漂移甚至建图崩溃。对此,提出一种基于因子图优化的多传感器信息紧耦合算法(tightly-coupled lidar-visual-inertial odometry via smoothing,mapping and DBSCAN,LVI-SMAD),将前端点云和视觉信息联合的聚类结果作为因子图优化约束,以一种较低帧的约束形式加入到较高帧的点云地图输出中,加强了点云与视觉信息的紧耦合,解决了激光雷达与相机间信息匹配错误的问题,同时将该约束作为某一传感器信息不可信时的约束补充,减小了传感器信息不稳定情况下的定位漂移,提高了算法一致性。实验证明,在低坡度长跨度的工作环境中,LVI-SMAD与LVI-SAM对比,绝对轨迹误差降低了39.90%,与LIO-SAM对比降低了63.09%;在高坡度工作环境中,与LVI-SAM对比,绝对轨迹误差减少41.08%,与LIO-SAM对比减少64.87%,证明了算法的有效性与可行性。展开更多
利用江苏南京2009-2012年天气雷达数据结合地面自动站风场资料分析江苏沿江地区阵风锋变化特征、阵风锋弧长与移速关系,及其在雷达反射率因子图像中呈现的总体、局部特征,详细分析了三种窄带回波回波带反射率因子分布特征。通过设计反...利用江苏南京2009-2012年天气雷达数据结合地面自动站风场资料分析江苏沿江地区阵风锋变化特征、阵风锋弧长与移速关系,及其在雷达反射率因子图像中呈现的总体、局部特征,详细分析了三种窄带回波回波带反射率因子分布特征。通过设计反映回波带平坦性的计算方法实现定量分析窄带回波分布异同功能。根据回波带径向波形特征判断径向波形的波宽、波峰个数、波峰阈值和波形双边梯度等特性,实现阵风锋径向波段识别。在对反射率因子图像预处理基础上,结合回波平坦性测试方法和阵风锋径向波形识别算法达到自动识别阵风锋回波的目的。识别效果表明:回波带平均值>5 d BZ的独立阵风锋回波识别准确率达87%以上,回波带平均值>10 d BZ的混合型阵风锋回波识别准确率达89%以上。对弱阵风锋识别成功率仍较低。展开更多
文摘By Doppler weather radar and satellite cloud data,the Meiyu front rainstorm which occurred in the north of Huaihe River and the north region of Yangtze-Huaihe was analyzed during July 1-2,2006.The characteristics of infrared satellite cloud,radar basic reflectivity and Doppler radial velocity were discussed in this rainstorm,such as the echo intensity,the characteristics of velocity field and so on.The reasons for this rainstorm were analyzed by these characteristics.The combination of radar and satellite data in the role of forecasting Meiyu front rainstorm was studied.
文摘多传感器建图与定位SLAM系统(simultaneous localization and mapping)在室外长距离跨度环境中,由于各传感器信息融合不正确、特征匹配错误,或传感器状态信息不可信,导致建图精度不足,轨迹漂移甚至建图崩溃。对此,提出一种基于因子图优化的多传感器信息紧耦合算法(tightly-coupled lidar-visual-inertial odometry via smoothing,mapping and DBSCAN,LVI-SMAD),将前端点云和视觉信息联合的聚类结果作为因子图优化约束,以一种较低帧的约束形式加入到较高帧的点云地图输出中,加强了点云与视觉信息的紧耦合,解决了激光雷达与相机间信息匹配错误的问题,同时将该约束作为某一传感器信息不可信时的约束补充,减小了传感器信息不稳定情况下的定位漂移,提高了算法一致性。实验证明,在低坡度长跨度的工作环境中,LVI-SMAD与LVI-SAM对比,绝对轨迹误差降低了39.90%,与LIO-SAM对比降低了63.09%;在高坡度工作环境中,与LVI-SAM对比,绝对轨迹误差减少41.08%,与LIO-SAM对比减少64.87%,证明了算法的有效性与可行性。
文摘利用江苏南京2009-2012年天气雷达数据结合地面自动站风场资料分析江苏沿江地区阵风锋变化特征、阵风锋弧长与移速关系,及其在雷达反射率因子图像中呈现的总体、局部特征,详细分析了三种窄带回波回波带反射率因子分布特征。通过设计反映回波带平坦性的计算方法实现定量分析窄带回波分布异同功能。根据回波带径向波形特征判断径向波形的波宽、波峰个数、波峰阈值和波形双边梯度等特性,实现阵风锋径向波段识别。在对反射率因子图像预处理基础上,结合回波平坦性测试方法和阵风锋径向波形识别算法达到自动识别阵风锋回波的目的。识别效果表明:回波带平均值>5 d BZ的独立阵风锋回波识别准确率达87%以上,回波带平均值>10 d BZ的混合型阵风锋回波识别准确率达89%以上。对弱阵风锋识别成功率仍较低。