Computation of the Travelling Salesman Problem by a Shrinking Blob

TL;DR

This paper introduces an unconventional, material-based computational approach to approximate solutions for the Euclidean TSP by shrinking a virtual blob that morphologically adapts to city locations, offering a simple and spatially intuitive heuristic.

Contribution

The paper presents a novel, morphology-driven method for approximating the TSP using a shrinking virtual material, differing from traditional algorithms by leveraging emergent shape adaptation.

Findings

The method achieves tours within approximately 4-9% of optimal length on average.

It demonstrates a simple, spatially intuitive heuristic for TSP approximation.

The approach reveals insights into proximity graphs and human performance models on TSP.

Abstract

The Travelling Salesman Problem (TSP) is a well known and challenging combinatorial optimisation problem. Its computational intractability has attracted a number of heuristic approaches to generate satisfactory, if not optimal, candidate solutions. In this paper we demonstrate a simple unconventional computation method to approximate the Euclidean TSP using a virtual material approach. The morphological adaptation behaviour of the material emerges from the low-level interactions of a population of particles moving within a diffusive lattice. A `blob' of this material is placed over a set of data points projected into the lattice, representing TSP city locations, and the blob is reduced in size over time. As the blob shrinks it morphologically adapts to the configuration of the cities. The shrinkage process automatically stops when the blob no longer completely covers all cities. By manually tracing the perimeter of the blob a path between cities is elicited corresponding to a TSP tour. Over 6 runs on 20 randomly generated datasets of 20 cities this simple and unguided method found tours with a mean best tour length of 1.04, mean average tour length of 1.07 and mean worst tour length of 1.09 when expressed as a fraction of the minimal tour computed by an exact TSP solver. We examine the insertion mechanism by which the blob constructs a tour, note some properties and limitations of its performance, and discuss the relationship between the blob TSP and proximity graphs which group points on the plane. The method is notable for its simplicity and the spatially represented mechanical mode of its operation. We discuss similarities between this method and previously suggested models of human performance on the TSP and suggest possibilities for further improvement.