Uncertainty-Based Vibration/Gyro Composite Planetary Terrain Mapping.

Area of Science:

• Robotics
• Geospatial analysis
• Sensor fusion

Background:

• Planetary rover missions depend on precise terrain perception.
• Vision and LIDAR-based terrain mapping are susceptible to environmental interference like dust and illumination.
• Existing methods struggle with uncertainty in terrain detection.

Purpose of the Study:

• To propose a robust terrain estimation method for planetary rovers.
• To address environmental uncertainties affecting sensor-based terrain mapping.
• To enhance the accuracy and reliability of terrain reconstruction.

Main Methods:

• Developed a vibration/gyro coupled terrain estimation algorithm.
• Incorporated multipoint ranging information.
• Derived a terrain update model considering measurement and motion uncertainties.
• Validated the method using a Clearpath Jackal unmanned vehicle in ROS simulation, indoor Optitrack, and outdoor soil environments.

Main Results:

• The proposed algorithm demonstrates strong terrain reconstruction capabilities.
• Achieved high accuracy, with reconstruction errors within 1 cm (simulation), 2 cm (indoor), and 6 cm (outdoor).
• The method effectively handles environmental uncertainties.

Conclusions:

• The vibration/gyro coupled method provides accurate terrain estimation for planetary rovers.
• The algorithm is robust across diverse testing environments.
• This approach enhances the feasibility of autonomous planetary exploration.