Articulated Robot Motion Planning Using Ant Colony Optimisation

A new approach to robot motion planning is proposed by applying ant colony optimization (ACO) with the probabilistic roadmap planner (PRM). The aim of this approach is to apply ACO to 3-dimensional robot motion planning which is complicated when involving mobile 6-dof or multiple articulated robots. An ant colony robot motion planning (ACRMP) method is proposed that has the benefit of collective behaviour of ants foraging from a nest to a food source. A number of artificial ants are released from the nest (start configuration) and begin to forage (search) towards the food (goal configuration). During the foraging process, a 1-TREE (uni-directional) searching strategy is applied in order to establish any possible connection from the nest to goal. Results from preliminary tests show that the ACRMP is capable of reducing the intermediate configuration between the Initial and goal configuration in an acceptable running time

I. Introduction

This paper describes a novel application of swarm intelligence to robotic arm manipulator motion planning. The probabilistic roadmap (PRM) is among the most efficient methods for planning robot motion. A PRM is a discrete representation of a continuous configuration space (C-space) generated by randomly sampling the free configurations of the C-space and connecting the points into a graph. The Single query Bi-directional with Lazy collision checking PRM (SBL-PRM) [11], [13] is a variation of the PRM which, instead of pre-computing the roadmap, uses two input query configurations as seeds to explore the space. It explores the robot's free space by concurrently building a roadmap made of two trees rooted at the query configurations (bi-directional). It delays collision tests along the edges of the roadmap until they are needed (lazy collision checking). Ant Colony Optimisation (ACO) is a swarm intelligence approach to solving optimisation problems. ACO has been successfully used to optimise the travelling salesman problem (TSP) and the quadratic assignment problem (QAP) [5]. To date, the applications of swarm intelligence in robotics have been primarily with collections of robots or with sets of transfer functions in factory production lines to solve one major task. References [16], [17] managed to solve the robot path planning for 2-dimensional space which involved a mobile robot of 2 to 3 degrees of freedom (dof). Applying ACO to a robot motion planner in which a mobile robot has 6-dof or more in 3-dimensional space, it is not straightforward. This paper introduces a new approach by applying ACO to robot motion planning, of manipulator arm type robots. It focuses on applying ant colony behaviour to search the robot's C-space. Preliminary experiments have been undertaken using SBL-PRM multi-goal motion planning as a benchmark for this new algorithm. Our aim is to get a faster planner and reduce the number of intermediate configurations between the two query configurations start and goal).