Research And Simulation Of Multi-UAV Mission Planning Algorithms In Dynamic Environment

UAVs are widely used to perform various dangerous tasks because of their small size and strong survivability.With the increasing demand for autonomy and collaboration in UAV mission systems,more and more scholars are focusing on improving the efficiency and performance of multi-UAV mission planning.The research directions mainly cover task allocation technology,trajectory planning technology,obstacle avoidance strategy,and multi-UAV collaboration technology.The research in this paper starts from assisting multi-UAV plan out strategies for performing detection,reconnaissance,and cruise missions in a complex environment of mountains.Through the research of multi-UAV task allocation algorithms and trajectory planning algorithms,this paper designs and implements a multi-UAV mission planner which can model the actual three-dimensional terrain and environmental information,obtain the optimal execution plan through task allocation algorithm and give the safe flyable track by the trajectory planning algorithm according to the information of multiple tasks and multiple different types of UAVs.This paper divides the research on multi-UAV mission planning in a dynamic environment into two stages.First,this paper conducts the research on task allocation,including model establishment,algorithm proposal and simulation performance test.Second,the research on trajectory planning includes the establishment of trajectory evaluation indicators,the fusion and improvement of algorithms and simulation test.The main content of this paper has the following three aspects:1.Aiming at the solution of multi-UAV task allocation,this paper analyzes the shortcomings of the existing algorithms and proposes a hybrid algorithm based on the idea of particle swarms for the beetle search,which improves the flaws of the particle swarm algorithm while improving the optimization speed and accuracy of the algorithm,and by adding a redistribution mechanism of contract network algorithms based on the market mechanism,which can quickly responds to unexpected situations during task execution.The final optimization stability of the task allocation algorithm reaches 86%,and the average optimization benefit value has increased by 23%.2.Aiming at the problem of multi-UAV trajectory planning,this paper analyzes the existing problems of the existing algorithm,and proposes an improved ant colony algorithm fused with particle swarms to improve the early convergence speed and final optimization accuracy of the algorithm,and enables the UAV to dynamically avoid obstacles during the execution of the mission by adding a Dubins curve method based on the potential function.The final algorithm reduces the cost of the track planned in the same time by about 10%.3.Design a mission planner based on Matlab,with GUI interface as front-end display,multi-UAV task allocation and trajectory planning algorithm as back-end data processing.Then using this mission planner to perform multi-UAV mission planning simulation in complex dynamic environment,and give the task allocation plan and the flight path of each UAV to perform the tasks.In summary,this paper has conducted research on algorithm theory to the integration and improvement of mission planning algorithm,and finally the design and implementation of multi-UAV mission planner and simulation verification,which proved the scientificity,feasibility and effectiveness of the research algorithm proposed in this paper.