Abstract
Iterative improvement is an optimization technique that finds frequent application in heuristic optimization, but, to the best of our knowledge, has not yet been adopted in the automatic design of control software for robots. In this work, we investigate iterative improvement in the context of the automatic modular design of control software for robot swarms. In particular, we investigate the optimization of two control architectures: finite-state machines and behavior trees. Finite state machines are a common choice for the control architecture in swarm robotics whereas behavior trees have received less attention so far. We compare three different optimization techniques: iterative improvement, Iterated F-race, and a hybridization of Iterated F-race and iterative improvement. For reference, we include in our study also (i) a design method in which behavior trees are optimized via genetic programming and (ii) EvoStick, a yardstick implementation of the neuro-evolutionary swarm robotics approach. The results indicate that iterative improvement is a viable optimization algorithm in the automatic modular design of control software for robot swarms.
Highlights
In this article, we investigate the use of iterative improvement in the automatic modular design of control software for robot swarms
In the missions AGGREGATION WITH AMBIENT CUES (AAC), SHELTER WITH CONSTRAINT ACCESS (SCA) and FORAGING, we register lower performance for those behavior trees designed by GP than those designed by the other methods
Finite-state machines designed by those design methods based on iterative improvement outperform those finite-state machines that are designed by Chocolate, in the missions AAC, FORAGING, and GUIDED SHELTER
Summary
We investigate the use of iterative improvement in the automatic modular design of control software for robot swarms. A swarm is highly redundant, self-organized, and decentralized. This offers several advantages such as robustness towards failure of individual robots and scalability of the swarm (Dorigo, Birattari & Brambilla, 2014). The main challenge of swarm robotics is to conceive the control software of the individual robots in such a way that a desired collective behavior emerges. A general methodology for the manual design of robot swarms is still missing and existing approaches either operate under restrictive assumptions or are labor-intensive, time-consuming, error-prone, and nonreproducible (Brambilla et al, 2013; Francesca & Birattari, 2016)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have