Robot Path Planning by Traveling Salesman Problem with Circle Neighborhood: modeling, algorithm, and applications

TL;DR

This paper addresses the traveling salesman problem with circular neighborhoods, proposing a nonlinear model, linearization, and a two-phase solution approach, applied to robot path planning and UAV data collection.

Contribution

It introduces a novel TSP variant with circular neighborhoods, formulates a nonlinear model, linearizes it, and develops a two-phase solution method implemented in Cplex and Knitro.

Findings

Effective for small and medium instances

Applicable to robot path planning and UAV data collection

Demonstrates the model's practicality in real-world scenarios

Abstract

This study investigates the problem of traveling salesman problem with circular neighborhood (TSPCN). Instead of cities there are circles and each point on circle can be a potential visiting node. The problem is to find the minimum length Hamiltonian cycle connecting the circles. Among the various real life applications of the problem, this paper concentrates on robot path planning for the laser welding robot, and data collection in a wireless sensor network by unmanned aerial vehicles (UAVs). The TSPCN is formulated as a nonlinear model, the objective function is linearized, and as a solution procedure decomposed into a two-phase model. The Models are coded in Cplex and Knitro and solved for small and medium sized instances.