| In recent years,the research of Micro Air Vehicles(MAV)swarm has developed rapidly.They are often used in military or civilian fields to perform field reconnaissance and data acquisition tasks.This thesis focuses on the research of formation control,obstacle avoidance control and cooperation of MAVs swarm.The main research contents and contributions are as follows:First,the formation control of MAVs is studied.Aiming at the problem of low robustness and communication bottlenecks of traditional formation control algorithms,a formation control algorithm based on virtual control network is proposed.A virtual control point set is generated based on the target formation to form a virtual control network.By updating the control point parameters,formation and dynamic evolution control processes of various formations are realized.Furthermore,aiming at the problem of increasing the calculation amount caused by the expansion of the swarm scale in the algorithm,a virtual control point selection mechanism based on subregions was proposed.Compared with other formation control algorithms,this algorithm can form any desired formation,and is easy to implement large-scale swarm formation with better selfadaptability.Secondly,the obstacle avoidance control of MAVs formation is studied.Based on the idea of constructing a virtual agent on the surface of obstacles,combined with the artificial potential field method,the repulsive force is generated by the virtual agent to change the speed and direction of MAVs,and the coordinated control of MAVs formation and obstacle avoidance is realized.In this thesis,the traditional artificial potential field method is improved,and the repulsion distance is set as the elastic distance related to the speed and the relative angle between the obstacle and MAV,so as to increase the “braking” distance in extreme cases,and obstacle avoidance of MAV formation in different speed scenarios is achieved.After that,the obstacle avoidance performance of the fixed repulsion distance and the elastic repulsion distance are compared by software simulation.Finally,taking the regional reconnaissance and target hunting task as an example,the cooperation problem of MAVs swarm is studied,and an area reconnaissance and target hunting cooperation algorithm based on formation control is proposed.For the regional reconnaissance,a search mode is designed.Based on the anti-flocking behavior of solitary animals,a virtual agent is constructed to guide the MAV in the direction of maximizing area coverage.And,the position of virtual agent is updated based on the search history of the MAV,environmental conditions and the state information of the neighboring virtual agent.For target hunting,a hunting mode is designed.When the target is found,target round-up is implemented based on the formation control algorithm proposed in this thesis.And the switching mechanism between search mode and hunting mode is designed.Software simulation shows that the MAVs has good dynamic coverage and hunting effect. |
