Self-reconfiguration planning method for heterogeneous modular robots based on reinforcement learning algorithm

Abstract

The invention relates to a self-reconfiguration planning method for heterogeneous modular robots based on a reinforcement learning algorithm. First, an initial modular robot configuration and a target configuration are given, the total number of modules N is input, and the graph structure of the modules is established through an initialization process. ;Take the initial configuration as the root node, establish a Monte Carlo tree search, and stop the search when the termination condition is reached (the target configuration is found or n times of exploration have been performed). After each search, the planned path is given and the samples are saved; when the number of samples reaches a given value, the samples are input into the neural network for training, and the training parameters are updated; after the parameters are updated, the Monte Carlo search is performed again. The average number of steps in this search result should be less, and after each search is completed, update the planned path with the least number of steps.

Description

technical field [0001] The invention belongs to the field of artificial intelligence planning, and in particular relates to applying a reinforcement learning algorithm to optimize the planning of a modular robot in the process of autonomous deformation. Background technique [0002] With the wide application of robots in various fields, more and more robots are used in unstructured environments to complete operations. For unknown working environments and different work tasks, robots need to be able to adapt to the environment by changing their configuration. This kind of robot that can dynamically and autonomously change its configuration to meet the needs of the task is called a self-reconfigurable modular robot. The self-reconfigurable modular robot is composed of a series of unit modules with simple structure and different functions. In order to achieve better taskability and environmental adaptability, when the multi-module system is running, each module unit independen...

AI Technical Summary

Problems solved by technology

For robots with complex configurations and many modules, in order to ensure the solvability from the initial configuration to the target configuration, the existing algorithm will introduce an intermediate configuration as a transition, which will lead to too many modules moving and inefficiency

Images