Searching Mission Planning Of Multiple UAVs Under Uncertain Information

In the process of unmanned combat,the opportunity is fleeting,and the belligerents hide the true and show the false.If we master more comprehensive and accurate data,we will master the initiative of the battlefield.As an important part of unmanned weapon equipment,UAV plays an important role in target search and data collection,and will become the key force of data-driven unmanned combat in the future.However,the existence of target information and the difficulty of collecting data pose a severe challenge to the UAV search task planning.Considering the uncertain information of search area,this paper studies the search task planning of UAV.The main contributions are as follows:The task time allocation method of single UAV is studied.Different from the implementation of hydrological monitoring,urban aerial photography and other activities,in the battlefield information reconnaissance mission,factors such as target camouflage bring a lot of uncertainty to its data collection,so that the UAV must spend more time on the target area to obtain more reliable information.Considering the target value and uncertain information,this paper establishes a single UAV task time allocation model,and its optimization objective involves an exponential continuous function.To solve this problem,two kinds of nonlinear complementary functions with parameters are introduced,and the model is transformed into smooth equations by Karush-Kuhn-Tucker theory;Secondly,according to the characteristics of the problem,the solution method is reformed and innovated.Specifically,based on the nonlinear complementary function with parameters,an improved smooth Newton method is proposed;In the process of smoothing Newton iteration,by modifying the parameters’ value,the dynamic smoothing Newton method is proposed.In terms of theoretical innovation,this paper proves the semi-smooth,continuous differentiable,nonsingular Jacobian matrix and superlinear convergence of the improved smooth Newton method.Through a large number of numerical experiments,the effectiveness of the improved smooth Newton method and dynamic Newton method is verified.At the same time,the experimental results show that the dynamic smooth Newton method is more robust than the smooth Newton method when the solution variables increase.The task time allocation method under multi UAV cooperative path planning is studied.Aiming at the uncertain target in the process of reconnaissance and search,a task time allocation model under multi UAV cooperative path planning is established in this paper.The objective function of the model involves a nonlinear function in exponential form.The constraints of the model include two kinds of decision variables: flight path and mission time,which poses a challenge to the existing optimization methods.Based on the semi greedy idea,this paper proposes a semi greedy construction method of multiUAV flight path.On this basis,combined with the improved smooth Newton method,a new Mission t Ime allocatio N und Er coope Rative p Ath p Lanning(MINERAL)is proposed.This paper analyzes the complexity of the proposed algorithm,and shows that the time and space complexity of semi greedy construction,smooth Newton method and MINERAL algorithm obey polynomial series.In addition,the simulation experiments of multi-UAV information collection are carried out,and the results show the performance advantage of MINERAL algorithm in solving this problem.