Multi-robot task assignment method based on VDSOM algorithm

Abstract

The invention belongs to the field of multi-robot task assignment, and particularly relates to a multi-robot task assignment method based on a VDSOM algorithm. The method not only effectively realizestask assignment of multiple robots, but also effectively avoids obstacles in an environment, so that multi-robot task assignment is more practical and efficient. The method provided by the inventionmainly comprises the steps that a neural network is initialized; a winning neuron (robot) of an input neuron is found to determine that the task target point is performed by the robot; a neighborhoodfunction is designed to determine an input neuron affected by the winning neuron within a neighborhood; the winning neuron and neighboring neurons thereof are controlled to move to the position coordinate of the target; according to a vector direction method, the neurons move a certain distance to the target point on the premise of successfully avoiding obstacles in the environment; and a weight is modified to determine the next winning neuron.

Description

technical field [0001] The invention belongs to the field of multi-robot task distribution, in particular to a multi-robot task distribution method based on VDSOM algorithm. Background technique [0002] In recent years, with the intelligent development of modern industry and manufacturing technology, the demand for robot systems in human society is also increasing. Especially in production and life, there are more and more repetitive, complex, and dangerous tasks that challenge the limits of human beings. In order to reduce human labor intensity and increase work efficiency, scientists have been working since the 20th century. Research on how to use robots to replace human work. However, due to the increasing complexity of the robot's working environment and more and more fields that require robots, a single robot can no longer meet the needs of humans, and the analysis and design of robots has become increasingly difficult. Therefore, using multiple robots to form a mult...

AI Technical Summary

Problems solved by technology

[0006] (1) There are no obstacles in the two-dimensional workspace, so the multi-robot task allocation algorithm does not have practical benefits;

[0007] (2) The direction angle of the robot's movement is not considered. In the actual environment, the original SOM algorithm directly uses the deviation between the current position of the robot and the target position in the environment to plan the path of the robot, which may cause the situation of crossing obstacles. So that the robot cannot reach the corresponding task target point, so it cannot complete the task assignment smoothly

Images