Cooperative Searching Method Of Moving Targets For Multi-UAV Based On Stackelberg Equilibrium

With the development of Unmanned Aerial Vehicle(UAV)technology,it was increasingly applied to target search tasks in recent years.Such as the rescue for the loss of passenger aircraft,military target detection and other work.For tasks that are difficult for human search,UAV is more adaptable and can cover a wider range of areas.At the same time,it can reduce search costs and improve search efficiency.It will play an increasingly important role in civil search and military search.Since the moving targets in the search task often have certain escape ability,the UAV cannot adopt simple coverage search,and needs to fully consider the targets' path to make its own path decision.Aiming at the UAV's search task for enemy moving targets in the battlefield environment,the game relationship model between the two parties is established,and the corresponding multi-UAV collaborative search path planning strategy is obtained.This thesis mainly does the following work:Firstly,according to the search task considered in this thesis,the relationship model that both sides with long-distance detection capability is set up.The basic models such as search environment model,UAV model,map information model and communication model are established,which lays a foundation for the later research.Secondly,after considering the path selection factors(such as environmental search income,target discovery income,etc.)in the process of searching for UAV and mobile targets,the path selection income function is established,which is used to the utility function of each step.Due to the Nash equilibrium strategy requires that both players to make decision at the same time which is not apply to the game relationship in this paper,the leader-follower model of the stackelberg equilibrium strategy is just applicable.The stackelberg equilibrium strategy can find a combination of behaviors that both sides' benefits reach a relative maximum,we can obtain the optimal search path for the UAV.The UAV and the mobile target act as leaders and followers respectively.Combining the path selection benefit function with the stackelberg equilibrium theory,a multi-UAV single-step game search path strategy is obtained.Considering the shortcomings of the single-step game strategy may fall into the local optimum,the concept of multi-step predictive control is introduced,and the multi-UAV multi-step game search path strategy is obtained to realize the rolling path planning of the UAV.Considering that the two strategies used in the previous section are all based on weighted optimization,their importance is represented by giving them different weights.Therefore,a hierarchical fuzzy optimization method is proposed,which divides the target into a series of sub-optimization problems and solves them layer by layer.Finally,the performance judgment method of the search strategy is determined.After setting the system simulation parameters,the strategies proposed in this thesis are compared and analyzed by designing different experiments.Simulation verification and analysis show the effectiveness of the proposed method.