Research On Decision Making Method In Human-machine Mixed-initiative Mission Planning Under The Condition Of Squad Cooperative Combat

With the accelerating pace of global urbanization,the proportion of military conflicts in cities and nearby mountains around the world is higher and higher.In cities with complex terrain and numerous buildings,long-distance and large-scale firepower are easy to cause civilian accidental injury,and it is difficult to coordinate combatants and apply conventional equipment.In the urban environment,decentralized operations with squads as the basic combat unit can give full play to its advantages of strong mobility,good concealment and high coordination efficiency,combined with modern intelligent unmanned combat systems,such as small rotor UAVs that can monitor the external environment of buildings,light unmanned vehicles that can provide close fire support,micro unmanned systems that can accurately detect targets,etc.Form a new crew based manned unmanned cooperative combat mode to improve combat effectiveness to a greater extent.In the manned-unmanned cooperative combat mission of the squad,the battlefield situation is changing rapidly.The commander will face practical challenges such as single person commanding multiple unmanned systems,complex battlefield conditions,poor combat environment and urgent decision-making time.The manned unmanned cooperative combat itself also includes the characteristics and difficulties of heterogeneous platform,complex space,difficult coordination and diversified constraints.In order to solve the above challenges,this paper takes the cooperative operation of squads and multiple unmanned platforms as the basic style,faces the typical reconnaissance and strike mission scenarios,studies the decision-making method in human-machine mixed-initiative planning under the condition of cooperative operation,and assists squad members in operation,so as to obtain greater operational efficiency.Taking the problem decoupling,problem modeling and problem solving as the main line,this paper focuses on two key problems of decision-making in squad cooperative operation-cooperative operation mission decomposition and heterogeneous mission allocation.The main work is as follows:(1)A decision framework model for squad cooperative human-machine mixedinitiative planning is proposed.Aiming at the squad cooperative decision-making problem with multi-agent and multi coupling constraints,based on the human-machine mixed-initiative planning framework,a decision-making framework model in squad collaboration is proposed.It is divided and abstracted into two problems: squad cooperative combat mission decomposition and heterogeneous mission allocation.The decision-making process and characteristics of crew manned-unmanned cooperative operation are analyzed,and a battlefield model including battlefield environment,battlefield objectives and capability requirements is constructed.(2)The mission decomposition model and tactical timing optimization algorithm of squad cooperative operation based on PDDL are proposed.Aiming at the problem of how to quickly decompose mission objectives and generate and optimize subtask sequences under various constraints such as battlefield space,tactical coordination and time resources,the capabilities of various unmanned platforms are defined by using standardized modeling language PDDL to form a domain knowledge model base;Based on the dynamic environment and model base,the graph programming algorithm is used to decompose the mission objectives to get the task sequences of the unmanned platforms;Finally,according to the specified tactical cooperative constraints,the dynamic recursive algorithm is used to optimize the tactical timing of the task sequence,which can effectively improve the efficiency of timing optimization under the new constraints.(3)A spatiotemporal collaborative mission allocation method for heterogeneous multi unmanned platforms based on improved ant colony algorithm is proposed.For several unmanned platforms that can carry different loads and have different combat capabilities,how to allocate tasks under the constraints of meeting the capability requirements,action sequence requirements and time window requirements of each task point,the multiple unmanned platform mission allocation problem model is transformed into a multiple traveling salesman problem with special constraints.For the dynamic environment,the mission allocation results can be obtained according to the conditions such as load,ammunition,distance,speed and so on.The heuristic information and pheromone update of ant colony optimization are improved.The test shows that the improvement is effective.(4)The simulation experiments of typical squad cooperative urban combat mission scenarios are carried out.Aiming at the cooperative offensive operation between squads and unmanned platforms in the urban scene,the operation scenario and task scenario are designed.Based on the air-ground cooperative virtual reality combined with the simulated countermeasure system,the task decomposition and task sequence space-time cooperative task allocation under sqaud cooperation are completed under various constraints.It verifies the effectiveness and advanced nature of the decision-making method in the human-machine mixed initiative planning under the condition of squad cooperative operation.