Study On Cooperative Task Planning Method Of Multi-UAV System

The task planning problem for multiple unmanned aerial vehicle(UAV)system exists in many fields,such as cooperative operation,data collection and disaster rescue.The purpose of this problem is to control the UAV system to execute the assigned tasks efficiently with the fewest resources in the complex and diverse task environment.The optimal task planning result should include the task assignment strategy,task execution sequence and navigation routes.The excellent task planning scheme is of great significance to improve the cooperation efficiency of UAV system.First,aiming at the task assignment problem of heterogeneous UAV system in static environment,the task assignment model with multiple constraints such as limited resources,task priority and workload balance is formulated.Then,in view of the shortage of solving speed and precision of intelligent algorithm,the concept of fuzzy elite degree is proposed to optimize and balance the transmission of good genes and the strength of population variation during the operation of algorithm.Based on the concept,an improved genetic algorithm is proposed to solve the constructed task assignment model.In the proposed improved genetic algorithm,two unlock methods are presented to solve the deadlock problem in the random optimization process.Second,the discrete adaptive search whale optimization algorithm is proposed to solve the problem of multi-heterogeneous UAV cooperative task planning in the static environment with multiple obstacles.First,the obstacle-avoiding distance estimation method is proposed to provide precise task scenario information and reduce the computational complexity.Then,the crossover-based updating method is presented to enable the algorithm to solve discrete problems such as task assignment.In addition,the search intensity adaptive mechanism is presented to optimize the exploration and exploitation intensities in the solution space reasonably,thereby solving the problem of insufficient exploration ability in the later stage of the algorithm.Further,the novel updating mechanisms are proposed to enhance the exploration ability of the algorithm.Finally,aiming at the problem of multi-UAV system cooperative task planning in the dynamic environment,the enhanced SOM neural network algorithm is proposed to improve the rapid adaptability of UAV system.However,the accuracy of the algorithm needs to be improved.To resolve the issues of low resolution accuracy of SOM algorithm,the quality parameter is proposed to evaluate the comprehensive value of tasks,thereby adopting the environmental information more effectively.Then,a new weight updating mechanism,which synthetically considers more abundant information such as environment and algorithm optimization is proposed.In addition,a neuron deletion mechanism is presented to further improve the fast response ability of the algorithm in dynamic environment.To sum up,this paper respectively propose the improved heuristic algorithm and SOM algorithm for the multi-UAV system cooperative task planning in static environment and dynamic environment.The proposed algorithms can improve the task execution ability of UAV system in a variety of operation scenarios.Simulation results demonstrate the effectiveness and superiority of the proposed algorithms.