Application Of Improved Immune Algorithm In Multi-robot Formation Control

Multi-robot collaboration can replace people to complete some complicated and dangerous tasks,so multi-robot coordination and cooperation as a branch of the robot research field has received more and more attention.As a fundamental problem of multi-robot coordination and cooperation,multi-robot formation plays an important role in the military and civil field.Therefore,it is of great significance to research related technology of multi-robot formation control.This paper starts from whether there are obstacles in the environment,and divides the formation problem into two situations: formation without environmental constraints and formation with environmental constraints,and research on multi-robot formation in these two scenarios.Firstly,aiming at the multi-robot formation problem without environmental constraints,proposes a formation control method of Lyapunov direct method.First based on the wheeled robot kinematics model with non-holonomic constraints,this paper uses the multi-robot formation structure of the pilot follower to introduce the virtual robot and introduce the multi-robot formation problem.In order to convert multi-robot formation problem into the problem of following robot tracking the virtual robot,the pose error system model of the following robot and the virtual robot is established.The controller is constructed by constructing the Lyapunov function and the stability is deduced by Lyapunov stability theory.Finally,according to the proposed tracking control algorithm,different formation follow-up mechanisms are established.The formation tasks of different formations are realized and the effectiveness of the proposed controller is verified through simulation experiments.Secondly,in view of the environmental constraints,multi-robots need to plan a collision-free path as a navigation path in advance.This paper proposes a path planning algorithm based on improved immune cloning.There are some questions in the process of path planning through traditional immune cloning algorithm: 1)Initialization path population is so slow that it affects algorithm optimization efficiency;2)Mutation operator is complex for all cloned antibody samples;3)Inhibition operator only with affinity Related to the degree,the population diversity is poor.In this paper,the path planning algorithm of the improved immune cloning algorithm is proposed: 1)using the improved artificial potential field method to plan several collision-free paths as the initial path population of the algorithm;2)proposing adaptive windowing mutation operator,only partial antibody samples The mutation is carried out;3)the antibody concentration is introduced into the inhibition operator,and the affinity value of the antibody and the concentration value together determine the probability of the antibody being inhibited.The simulation experiment realizes the effect of the path planning through the improved immune cloning algorithm under different map scales,and verified the effectiveness of the improved immune cloning algorithm for path planning.Finally,the problem of being able to avoid obstacles while following the robot in a complex environment while maintaining the formation,proposes a method for optimizing behavioral parameters using an improved immune cloning algorithm.In this paper,we first decompose it into several sub-behaviors of the following robots.Then we use vector synthesis to synthesize the behavior of the robot.Finally,we use the improved immune cloning algorithm to optimize the behavior parameters of the robot.The simulation experiment realizes that after using the improved immune cloning algorithm to optimize the behavior parameters,the multi-robot can reach the target point with a certain formation without collision in the complex environment,and compares with the behavioral parameters based on the empirical method.The result verifies the effectiveness of the proposed algorithm.