Optimization Of Modular Robot Configuration Based On Genetic Algorithm

Reconfigurable modular robot is composed of a number of modules with the same interface,according to different tasks to select different modules assembled into different performance of the robot.Because of its short design cycle,low cost,good adaptability to different tasks and environment,it has become a hotspot in robot research.In this paper,the module is divided into the base module,the joint module,the link module and the end actuator module.Then establish the coordinate system at the interface of each module,and find the correlation matrix of each module entrance coordinate system and exit coordinate system.From the base to the end of the actuator,the port near the base is the inlet,and the port near the end effector is the outlet.The expression of the robot structure is made by using the method of correlation matrix,which makes the matrix and the robot structure correspond.The positive and negative solutions to the kinematics are studied so that positive and negative solutions can be obtained quickly for a given arbitrary configuration.For the positive solution of kinematics,the kinematic equation is obtained by multiplying the transformation matrix corresponding to each module,and the kinematic positive solution can be obtained by substituting the joint angle into the kinematic equation.Genetic algorithm is used to solve the inverse solution,and the kinematics simulation of the robot is carried out by using Solidworks to verify the correctness of the kinematics equation.For the modular robots face the task to find the right module assembled into a robot.Complete the corresponding task.The performance index of the robot is analyzed,and the accessibility index,the work space index and the dexterity index are mainly studied.On this basis,the non-dominate sorting genetic algorithm(NSGA-II),which introduces the density estimation operator and the crowding comparison operator,is used to optimize the robot configuration in the face of specific tasks.The paper analyzes the specific handling tasks,and optimizes the correctness of the robot by optimizing the freedom index,accessibility index,motion transmission index and work space index of the robot.