Cooperative Mission Decomposition And Scheduling Of Man-Unmanned System Under LTL Specification

Man-Unmanned cooperation is the most practical and feasible combat mode for UAVs in antagonistic environment,which is also the most effective way to generate new efficient operational system.Under the condition of complex mission and low Man-Unmanned ratio,traditional aerial command and control method can't meet the needs of cooperative mission.Meanwhile,Temporal-logical-based languages provide a formal and expressive way to specify high-level complex cooperative tasks for multiple UAVs.Based on the LTL specification and automaton theory,this dissertation systematically studied the approaches of cooperative mission planning and scheduling for single and multiple UAVs.The main work and contributions are summarized as follows:(1)Based on the exiting work,a dynamic planning framework was proposed.After introducing the related knowledge of LTL,transition system and automaton,an existing planning method under LTL specification was briefly introduced,which leveraging ideas from automata-based model checking approaches.Based on this method,an improved framework was proposed to extend the synthesis to dynamic way,which enables replanning when environment changes unexpectedly.(2)An entire solution for dynamic planning of single UAV under LTL specification was proposed.To model this problem in formalized approach,predicate logic is used to formulate physical system,and PDDL languages are used to describe the action of UAV.This solution transferred the formalized description into transition system,thus proposed dynamic framework is utilized.In addition,to boost the searching speed,heuristic information was defined from the LTL specification and simulation result indicate 30% fast than non-information search.(3)The approach for cooperative mission scheduling of multi-UAVs is studied.Based on the formulation of single UAV,cooperative action mode for multiple UAVs was introduced and all actions are attached with time of duration,thus enabling the UAVs acting asynchronously.Making use of heuristic information from decomposed mission,an iterative extending and planning approach in infinite system state space is proposed to planning the schedule for cooperative mission.Simulation result demonstrated this method can markedly decrease the expanded state nodes and searching times,the total planning are finished within 1 second.