一种车辆4D毫米波雷达惯性里程计方法及计算机可读介质
By employing a 4D millimeter-wave radar inertial odometry method, combined with techniques such as multipath scattering removal, IMU data integration, and Kalman filtering, the problem of positioning difficulties for 4D millimeter-wave radar under adverse weather conditions was solved, achieving high-precision vehicle state estimation.
CN117452358BActive Publication Date: 2026-07-17WUHAN UNIV
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- WUHAN UNIV
- Filing Date
- 2023-03-13
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
4D millimeter-wave radar struggles to achieve stable, continuous, and high-precision data association under adverse weather conditions, leading to difficulties in vehicle positioning and mapping.
Method used
A high-precision vehicle state estimation model is constructed by employing a 4D millimeter-wave radar inertial odometry method, through steps such as multipath scattering and speckle noise removal, IMU data integration, Kalman filtering, point cloud matching, and closed-loop constraints.
Benefits of technology
It improves the matching accuracy and robustness of sparse radar point clouds, reduces the impact of noise on positioning results, and enhances positioning accuracy under adverse weather conditions.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN117452358B_ABST
Abstract
本发明提出了一种车辆4D毫米波雷达惯性里程计方法及计算机可读介质。通过松弛滤波剔除显著的雷达空间噪声，基于选权迭代鲁棒估计法进行单帧测速并剔除动态噪声；通过机械编排进行IMU观测积分得到车辆预测状态；基于单帧测速和预测速度构建速度约束初步更新车辆状态；基于之前点云帧构建局部地图，并搜索下一时刻每个radar点的局部地图n近邻点，计算分布到多分布加权距离作为点云到局部地图的匹配约束，基于迭代滤波进一步更新得到精确车辆状态；通过4D radar闭环检测和GICP更新所有时刻的车辆状态，降低定位漂移。本发明削弱了各类雷达噪声对定位结果的影响，并有效提高了稀疏雷达点云的匹配精度和鲁棒性。
Need to check novelty before this filing date? Find Prior Art