Robust moving horizon planning for multi-vehicles area coverage in uncertain environment using mixed-integer-programming

The increased use of multi-vehicles raises concerns about safety and economic aspects in several applications. Therefore, this work proposes a moving horizon planning algorithm for covering unexplored regions using multi-vehicles in uncertain/dynamic environments. The proposed algorithm enables the vehicles to adapt online to changes in the environment despite wind disturbances and vehicle uncertainties. The proposed planning allows systematic consideration of vehicle dynamics and constraints, e.g., obstacle avoidance, for optimizing a performance index, e.g., uncovered area and energy consumption. The algorithm robustness is demonstrated through theoretical investigations and numerical simulations in various uncertain scenarios using different planning architectures, e.g., centralized, decentralized, and distributed. The distributed planning approach achieves the best performance in terms of the covering rate, robustness, and computation time.