A quadruped robot positioning and mapping method and device in a complex terrain environment, a terminal and a medium

By constructing a state estimation observation model and a confidence-aware planar constraint update mechanism, the instability of state estimation and elevation drift in the localization and mapping of quadruped robots under complex terrain are solved, achieving high-precision 3D mapping and robust pose estimation.

CN121977535BActive Publication Date: 2026-07-14SHENZHEN UNIV
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SHENZHEN UNIV
Filing Date
2026-04-03
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

Existing technologies for localization and mapping of quadruped robots in complex terrain environments suffer from state estimation instability, sensor data distortion, and elevation drift, making it particularly difficult to achieve high-precision 3D mapping in highly dynamic and unstructured environments.

Method used

By constructing a state estimation observation model and combining inertial measurement data and leg motor data, motion state estimation is performed. Based on a confidence-aware planar constraint update mechanism, the state estimation process of the quadruped robot is optimized. Motion compensation and point cloud registration are performed by combining LiDAR point cloud data, and a three-dimensional voxel map is output.

Benefits of technology

It improves the stability of pose estimation and the accuracy of 3D mapping for quadruped robots in complex terrain, and can provide reliable short-time pose estimation and high-quality 3D mapping under irregular motion conditions, thus enhancing the robustness and consistency of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121977535B_ABST
    Figure CN121977535B_ABST
Patent Text Reader

Abstract

The application discloses a kind of complex terrain environment under quadruped robot positioning and mapping method, device, terminal and medium, method includes: obtaining laser radar point cloud, inertial measurement data, leg motor data of quadruped robot, constructs the state estimation observation model of quadruped robot, and realizes the estimation of the motion state of quadruped robot, obtains state estimation result;Based on state estimation result point cloud motion compensation, and based on state estimation result, the global map of laser radar point cloud and adaptive voxel structure after compensation, realize point cloud registration and point cloud state estimation, output three-dimensional voxel map;Based on confidence perception plane constraint foot end update mechanism, three-dimensional voxel map is used as constraint update foot end elevation, to optimize the subsequent state estimation process of quadruped robot.The application can effectively solve the problems of positioning drift, matching loss, mapping ambiguity and registration failure caused by flexible motion state of quadruped robot under complex terrain.
Need to check novelty before this filing date? Find Prior Art