A Mixed-Logical-Dynamical Model for Multi-Robot Lunar Construction Task Allocation

This paper proposes a unified framework for allocating construction tasks among multiple heterogeneous robots on the lunar surface. Discrete task and charging decisions, continuous battery and task progress dynamics, and task allocation logic rules are integrated into a hybrid system, which is then formulated as a Mixed Logical Dynamical (MLD) system. The task allocation problem is cast as a constrained Mixed-Integer Quadratic Program (MIQP) problem addressed via Model Predictive Control (MPC). While enabling real-time adaptability, the MLD constraints consider task precedence, robot cooperation bounds, and energy limits. Simulation case studies in lunar conditions confirm the effectiveness and robustness of the approach. Future work will focus on decision interactions among multiple robots based on mixed-integer potential games to further enhance overall system performance.