The Study Of The Flocking And Obstacle Avoidance Of Multiagent Systems With Differences

The design of individuals in multiagent systems is limited by cost constraints and design difficulties.Most studies have used the same individual model and ignored the intra-group variability.In this master’s thesis,the effects of individual differences,group differences,and mutants on multi-agent systems in flocking and obstacle avoidance movements are investigated.The differences are mainly in the perceptual radius,inter-individual forces,obstacle avoidance ability,and target tracking ability.We designed a smooth and bounded hybrid potential function with indefinite parameters.This function not only satisfies the consistency control requirements of the three previously mentioned systems,but it also applies to the general cluster system without individual differences.As a result of the action of this function,the system has the advantages of rapid swarming and constant system connectivity during motion.The simulation is carried out in 2D and 3D space for the above-mentioned different cases.Simulations show how the differences within the system affect the swarming and obstacle avoidance trajectories of the group and the stratification phenomenon.The results confirm the theory’s validity.The key contributions of this study,as compared to earlier research,are as follows:(1)The majority of previous coherent control experiments generate a sophisticated complete scenario energy field.In this master’s thesis,the interaction mechanism of micro-scopic particles is used to investigate the force situation of the individual agent.To that end,we created the movement model of the individual agent and investigated thoroughly their swarm movements.(2)We considered the different differences that exist between individuals and between groups.It enables individuals or groups with varying abilities to display diverse motion effects in consistent motion.This is reflected in such things as interaction forces(attractive and repulsive force),perceptual capacity(communication),and behavioral ability(target tracking,speed matching,obstacle avoidance).(3)The movement of the mutant agent with various potentials and radii is investigated.The special individual with "great influence" on the surrounding individuals can act as a true leader,and followers can move without needing to obtain information about the target.(4)In this thesis,we consider the topological connectivity situation in swarm motion.By constructing a specific potential energy function that allows the interacting agents to generate a large enough attraction at the interaction edge,the system has good connectivity retention capability.