Research On Multi-UAVs Mission Planning Algorithm For Dynamic Battlefield Environment

Multi-UAV mission planning technology is one of the indispensable core technologies for UAV operations.It can be divided into two parts: upper-level task assignment and lower-level trajectory planning.The battlefield situation in modern warfare is more complex and changeable,which puts forward higher requirements for the multi-UAV mission planning system: In terms of task assignment,the mapping relationship between UAVs and tasks is becoming more and more complex,and it is necessary to consider the task Execution sequence,execution time window,UAV range and speed constraints,which put forward higher requirements on the problem modeling of task assignment and the performance of task assignment algorithm; The planned trajectory puts forward higher requirements in terms of safety and flyability; in terms of task redistribution,there are various emergencies in the battlefield,which may partially invalidate the original planning scheme,requiring multi-UAV missions The planning system responds and handles emergencies quickly,taking into account optimality while ensuring real-time performance.Therefore,in order to meet the increasing demand for multi-UAV in modern warfare,it is of great significance to study multi-UAV mission planning technology.In this thesis,based on the research background of multiple UAVs performing ground strike missions in mountainous environments,the existing theories and methods of task assignment and track planning are analyzed and corresponding improvements are made,and a task reassignment processing scheme is designed.The proposed improved genetic algorithm task assignment and improved particle swarm algorithm track planning form a complete multi-UAV mission planning scheme.Finally,the effectiveness of the scheme proposed in this thesis is proved by the simulation experiments in 3D static and 3D dynamic environments.The main research contents are as follows:In terms of task allocation,through the analysis of existing algorithms,the thesis designs task allocation test experiments to select the most potential genetic algorithm.In response to problems such as the high pressure of the selection operation in the later stage of genetic algorithms,the complex convergence speed of single-generation evolutionary computation,and the insufficient ability to jump out of local optimal traps,the thesis makes improvements to the three basic operations of selection,crossover,and mutation.The simulation comparison experiments on four typical optimization test functions have proved the effectiveness of the improvements.In response to the difficulties in designing genetic coding for genetic algorithms and the tendency of crossover and mutation operations to make genes invalid,the thesis adds the trajectory cost to the genetic coding and designs corresponding coding rules.For the crossover and mutation operations,specific processes and conflict resolution methods are also designed to solve the application problems of genetic algorithms in task allocation.In terms of trajectory planning,in response to problems such as the tendency of particle swarm optimization to fall into premature traps when applied to trajectory planning problems and the inability of the planned trajectories to adapt to the drone’s maneuvering constraints,the thesis adds inertia weight to the particle swarm optimization algorithm to improve its ability to resist premature convergence.The thesis also makes improvements to the trajectory coding method corresponding to the particles,mapping the solution space from the Cartesian coordinate system to the spherical coordinate system,and designing relevant constraints to include the drone’s pitch angle and azimuth angle in it,so that the planned trajectories have better safety and maneuverability.This operation also reduces the search space and speeds up the convergence speed of the algorithm.In terms of handling/task reassignment of dynamic emergencies,in view of the problem that the existing algorithms and processing schemes cannot balance real-time performance and optimality,this thesis discusses the problems of UAV disconnection/crash,new tasks,and new threat sources.A detailed problem-solving analysis has been carried out in three emergencies,and a task redistribution scheme that can search for a better solution under the premise of ensuring real-time performance is proposed.