Research On Docking And Combination Strategy Of Reconfigurable Modular Robot

Nowadays,modular robots have become a popular research object in the field of robotics.By changing the connection and combination of multiple modules,a group of modular robots can form various configurations for adapting complex environments or performing different tasks.For example,when several modular robots encounter obstacles exceeding the size of a single module,they can combine several modules into a serpentine configuration to overcome the obstacles.This kind of configuration action can rely on self-reconfiguration.It also can be done with the help of external intervention.When a group of modular robots encounters materials that weigh more than the load of a single module,the cluster load capacity can be improved by combining into a matrix array configuration.In order to explore the problems of modular robots in practical applications,this paper proposed the design principle of the control and operating system of reconfigurable modular robots.Then the communication experiments of the docking mechanism and the optimization of the endpoint trajectory energy consumption based on Fourier function was conducted.At last,the approximate algorithm for reducing energy consumption of the multi-robot system is introduced.(1)Design principles for stable,low-latency control systems that support multitasking processes:Aiming at the working characteristics of reconfigurable modular robots,this paper analyzes the requirements of modular robots on the control system when performing tasks.Three design principles of reconfigurable modular robot control and operating system are summarized:real-time response,stable reliability,and multi-task execution capability.According to the above principles,the communication bus mode required by the robot is analyzed and selected and CANopen is chose as the communication mode for the data interaction between the docking mechanism and the control system.The control system carrier uses the ADLINK IB915AF-6600 embedded motherboard and the operating system uses ROS(Robot Operating System).The driver of the docking mechanism uses the ADM system driver and the joint motor uses a frameless torque motor as the torque source for driving mechanism motion.By optimizing the underlying data sending and receiving structure,the real-time communication is improved.The overall control system of the module is divided into upper and lower logical layers to reduce the coupling of data communication between different hardware for improving system stability.(2)Communication experiment of docking mechanism and optimization of endpoint trajectory energy consumption:Aiming at the energy consumption problem of the docking mechanism during the movement,this article follows the idea of assembling and debugging the physical prototype at first,then analyzing the theoretical simulation.After completing the adjustment of the PID parameters of the driver,the mechanism was modeled based on the Langer's equation.Next,a mechanism energy consumption model based on mechanical energy and motor copper loss is established.By comparing the traditional fifth-degree polynomial end trajectory equation with the Fourier equation to analyze the energy consumption of the mechanism,the current acceleration change trend of the docking mechanism during the movement is obtained.Then,the optimal energy motion curve based on the Fourier equation is formulated.A prototype docking experiment was performed in order to verify the reliability and stability of CANopen communication.(3)Optimization of energy consumption of multi-module robots:By analyzing the shortest path between two modular robots,an aggregation and reconstruction problem of multiple modular robots is proposed.The reconfiguration of a multi-module robot can be divided into two stages.This paper mainly analyzed the computational complexity of the energy consumption problem in the first stage of reconstruction and built the energy consumption model of the module platform.It is proved that the optimal convergence energy consumption problem of a multi-module robot is NP-complete.For solving this problem,an approximate algorithm based on genetic algorithm and greedy strategy is designed.Finally,simulation analysis and verification are performed.Previous studies on self-reconfiguration problems have focused on chained modular robots for decades.The innovation of this paper lies in introducing graph theory to optimize the energy consumption of the matrix modular-robot reconstruction and convergence path planning problem.An approximate algorithm based on the combination of genetic algorithm and greedy strategy is designed to solve this problem.It has enlightening significance for the subsequent reconstruction research of matrix modular robots.