Comparing SONN Types for Efficient Robot Motion Planning in the Configuration Space

TL;DR

This paper compares various Self-Organizing Neural Network (SONN) architectures for efficient robot motion planning in high-dimensional configuration spaces, extending previous models and validating trajectories in simulation.

Contribution

It introduces new SONN models for C-space reduction, adapting human motion data approaches to robot kinematics, and evaluates their effectiveness in trajectory planning.

Findings

Successful trajectory generation with multiple SONN models

Enhanced C-space reduction techniques for higher DOF robots

Validated robot trajectories in simulation environment

Abstract

Motion planning in the configuration space (C-space) induces benefits, such as smooth trajectories. It becomes more complex as the degrees of freedom (DOF) increase. This is due to the direct relation between the dimensionality of the search space and the DOF. Self-organizing neural networks (SONN) and their famous candidate, the Self-Organizing Map, have been proven to be useful tools for C-space reduction while preserving its underlying topology, as presented in [29]. In this work, we extend our previous study with additional models and adapt the approach from human motion data towards robots' kinematics. The evaluation includes the best performant models from [29] and three additional SONN architectures, representing the consequent continuation of this previous work. Generated Trajectories, planned with the different SONN models, were successfully tested in a robot simulation.