Intelligent multi-objective receding horizon control for UCAV mission planning

Abstract

Mission planning is the key capability of unmanned combat aerial vehicles (UCAV), especially when an UCAV performs penetration and attack mission in dynamic and adversarial environments. There are two challenges for automated mission planning. First, the environment and mission target maybe encounter dynamic change. Second, the several objectives are often conflicting and adjustable in different mission stages. Aim to conquer the difficulties, an intelligent and multi-objective receding horizon control framework is proposed. The framework integrates multi-objective optimization, receding horizon control, fuzzy inference system and expert knowledge together, which can online receding and adaptively adjust the objectives according to the environment and the state of the mission execution. For UCAV mission planning in adversarial environments, once the multi-objective optimization problem is solved at each sampling time and the non-inferior solutions belonging to the set of Pareto are obtained, the most satisfied one is selected by using the objectives weights inferred from the expert decision. Results are shown on typical examples of UCAV mission planning under dynamic environments, which illustrated the feasibility and applicability of the proposed method.