Mobile Robot Pose Estimation Using Point Cloud and IMU Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile robot attitude detection methods using inertial measurement units (IMUs) are inaccurate due to large error values, especially when encountering potholes or doorsills, affecting positional posture and obstacle sensing.
Innovation Solution
A posture estimation method combining point cloud data from a linear laser and IMU data to accurately determine the mobile robot's attitude, using least squares optimization to correct the robot's posture in the world coordinate system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If inertial measurement unit is used for attitude detection, then the robot can perform planar tasks, but the measurement precision is low due to large error values
Solution Approach 1:
The patent combines point cloud data from linear laser with attitude data from inertial measurement unit to perform joint optimization. This merging of multiple data sources compensates for the limitations of each individual sensor, achieving high-precision posture estimation that overcomes the low measurement precision of the IMU alone.
Solution Approach 2:
The patent introduces a work plane model as an intermediary between the IMU and the final posture estimation. The point cloud data serves as intermediate information that helps establish the work plane, which then acts as a reference for correcting the IMU measurements and achieving accurate attitude detection.
2Adaptability or versatility
If robot encounters potholes or doorsills, then the robot may bump and change attitude, but this causes incorrect obstacle marking and inaccurate positional posture
Solution Approach 1:
The patent dynamically adjusts the work plane model based on real-time point cloud data. When the robot encounters potholes or doorsills causing attitude changes, the system updates the work plane representation to reflect the new terrain conditions, maintaining accurate posture estimation despite the dynamic environmental changes.
Solution Approach 2:
The patent implements a feedback mechanism where the point cloud data continuously validates and corrects the work plane model. When the robot encounters unexpected terrain features, the point cloud provides feedback that triggers recalculation of the work plane, ensuring the posture estimation remains accurate even after bumps or attitude changes.
Data Source
Figure 1
Figure 2-a~2-b
Figure 3-a~3-b
AI summary
A pose estimation method and apparatus for a mobile robot, the mobile robot, and a storage medium. The method comprises: obtaining an environment three-dimensional point cloud when a mobile robot executes a planar task, and on the basis of a preset height threshold, determining a working plane point cloud according to the environment three-dimensional point cloud (S110); obtaining inertial measurement information of the mobile robot (S120); and according to the inertial measurement information and the working plane point cloud, determining the pose of the mobile robot in a world coordinate system (S130).