Simulation and Control of an Unmanned Lunar Rover at the Moon’s South Pole: Leveraging Challenging Illumination Conditions for Enhanced Exploration
- Paper ID
102536
- DOI
- author
- company
Middle East Technical University
- country
Türkiye
- year
2025
- abstract
This study explores the motion planning and control of a lunar rover designed for exploration at the Moon's south pole. The mission aims to explore permanently shadowed regions believed to contain water ice deposits, identified through orbital measurements in 2020. However, no direct imaging or physical sampling has been conducted in these regions. To achieve this, we propose deploying multiple micro-rovers for exploration, with a focus on validating Guidance, Navigation, and Control (GNC) algorithms in challenging lunar conditions. These conditions, including gravity and terrain mechanics, make Earth-based testing of these algorithms extremely difficult. To address this, we developed a high-fidelity simulation environment using Unreal Engine, which enables testing and optimization of GNC algorithms, including collaborative tasking among multiple rovers. The simulation focuses on optimizing key metrics, such as average power consumption and maximum speed to cover a 100-meter radius. This is achieved by iteratively optimizing the rover’s solar panel tilt angle for power efficiency. We also determine the maximum speed required to explore the area. Our model incorporates the rover’s weight, motor dynamics, terramechanics, and evolving internal parameters, ensuring vehicle performance and terrain interaction are considered. Mass is crucial in space missions, affecting power consumption, cost, and feasibility of space-grade components. While LIDAR sensors are effective, they present challenges in power, mass, and cost. Therefore, visual sensors are essential for micro-rovers under 5 kg. The extreme lighting at the Moon’s south pole, with the Sun at a low elevation for 14 days, creates near-constant backlighting, complicating visual sensing. To navigate, the rover lacks GPS and magnetometric measurements, relying instead on range data relative to the lander and a Sun sensor for solar direction. Our algorithm follows a Sun-referenced strategy to explore a 100-meter radius around the lander over the course of five lunar days. The area is split into five equal sectors, and the rover traverses the boundaries between adjacent sectors in a leaf-like pattern. This strategy results in significant area coverage, as each daily traversal extends the rover’s reach, ensuring efficient exploration over the total area. After each day’s exploration, the rover returns to almost the same radial position around the lander and proceeds to the next sector. This allows for the alignment of sector mappings, facilitating consistent tracking despite the challenges posed by the long, rotating shadows, which could potentially disrupt the Visual-Inertial Navigation (VIN) algorithms.