Transport network creation of target-oriented particle-based models of Physarum Polycephalum

Abstract

‘Physarum Polycephalum’ is a slime mold capable of solving the shortest path-problem. It tries to maximize its food intake while preserving minimal length of the network created between food points. Models of this organism have been constructed, which show the same properties. With certain parameters, these models can be used to approximate existing transport networks. Because real-life transport networks are solutions to the network design problem, an algorithm which approximates these networks well enough can also be used to design transport networks. We hope to increase the quality of network design by implementing the notion of targets in an existing Physarum-algorithm, creating a target-oriented particle-based model of Physarum Polycephalum. With this adaptation, particles will have a randomly assigned target, which is a food point. We show that this adaptation does improve the approximation of existing transport networks. Thus, we have created a model which is better at designing an efficient transport network than the regular Physarum-algorithm. We also provide some useful applications and further adaptations, which could further increase the quality of network design by Physarum-algorithms.