Research On The Method Of Multi-UAVs Cooperative Ground Mission Planning

With the continuous advancement of UAV technology,the status of multi-UAVs coordinated ground mission planning in modern warfare has become increasingly prominent,and the results of its planning will directly affect the overall combat effectiveness of UAVs.Multi-UAVs coordinated ground mission planning consists of two stages: the first stage is to assign a reasonable task to the UAV;the second stage is trajectory planning,that is,the UAV can safely reach the ground mission target.This thesis combines the research of a certain UAV simulation platform to carry out research on related issues and the main contents of the work are as follows:(1)In order to solve the problems of unbalanced resources,poor real-time performance,and low overall efficiency in the task allocation of multi-UAVs in the traditional contract network algorithm,this thesis combines the needs of multi-UAVs task allocation to improve the contract network algorithm.First of all,in order to obtain a realistic task allocation model,this thesis considers the battlefield environment,task requirements and UAV performance conditions when constructing the model.Secondly,in order to make the model have real-time monitoring capabilities,this dissertation adds two steps of "status update" and "heartbeat maintenance" on the basis of the traditional contract network algorithm.Compared with the traditional contract network method that can only perform offline task allocation,the improved method can carry out task allocation in both offline and real-time modes.Thirdly,in order to enable the model to achieve higher performance under multiple constraints,this dissertation introduces a multi-index constrained performance function.Specifically,this function takes into account the value of mission target revenue,flight range,and the degree of threat to the UAV.Finally,in order to improve the global load capacity of the model,this dissertation introduces a load balancing mechanism,and adds tasks such as task trading and task exchange during task allocation.The experimental results show that the model proposed in this dissertation has better global load balancing effect and shorter time overhead than traditional methods,and has the ability to allocate real-time tasks.(2)In order to solve the problems of poor real-time performance,slow convergence,and easy to fall into local optimum in the traditional PSO in solving the problem of multi-UAVs trajectory planning,this dissertation combines the actual trajectory planning requirements to improve the traditional PSO algorithm.First,this thesis uses the minimum surface method and hemispherical model to model the threat area and two-dimensional projection.Secondly,in order to solve the problem of slow convergence of the nonlinear function adaptive strategy,this dissertation designs a linearly declining adaptive strategy to approximate the optimal parameters of the PSO.In order to improve the efficiency of traditional PSO algorithm,this dissertation divides the PSO into multiple sub-groups with reference to the K-means algorithm,and refers to the niche algorithm from several sub-groups and select the best individuals to form a niche group.In order to solve the local optimization problem of the traditional PSO algorithm,we refer to the mechanism of the simulated annealing algorithm to jump out of the local optimum,and then propose an improved hybrid PSO algorithm based on the traditional PSO.Finally,our method will smooth the track based on the Dubins curve.Multi-scene experiments show that compared with traditional PSO optimization,hybrid PSO optimization can solve the global optimal solution at a faster convergence rate,which is suitable for solving trajectory planning problems in multiple scenarios,and can carry out real-time navigation track planning.(3)Aiming at the problem of the mismatch between the simulation platform and mission planning,this dissertation added corresponding interfaces in the process of integrating the simulation platform and mission planning based on mission planning requirements.Secondly,based on this platform,we implemented a bee colony demonstration simulation deduction project and applied the task allocation and trajectory planning methods proposed in this dissertation have achieved good results in mission deduction.Finally,the experimental results show that the planning effect and time of the platform can meet the actual task requirements,and it is suitable for coordinated ground combat tasks.Mission planning experiments have good engineering application value.