Comparing encodings for performance and phenotypic exploration in evolving modular robots

Veenstra, Frank; Hart, Emma; Buchanan, Edgar; Li, Wei; De Carlo, Matteo; Eiben, Agoston E.

Abstract

To investigate how encodings influence evolving the morphology and control of modular robots, we compared three encodings: a direct encoding and two generative encodings---a compositional pattern producing network (CPPN) and a Lindenmayer System (L-System). The evolutionary progression and final performance of the direct encoding and the L-System was significantly better than the CPPN. The generative encodings converge quicker than the direct encoding in terms of morphological and controller diversity.

Citation

Veenstra, F., Hart, E., Buchanan, E., Li, W., De Carlo, M., & Eiben, A. E. (2019, July). Comparing encodings for performance and phenotypic exploration in evolving modular robots. Presented at GECCO '19: Genetic and Evolutionary Computation Conference, Prague, Czech Republic

• Presentation Conference Type: Conference Paper (published)
• Conference Name: GECCO '19: Genetic and Evolutionary Computation Conference
• Start Date: Jul 13, 2019
• End Date: Jul 17, 2019
• Online Publication Date: Jul 13, 2019
• Publication Date: Jul 13, 2019
• Deposit Date: Aug 11, 2020
• Publisher: Association for Computing Machinery (ACM)
• Pages: 127-128
• Book Title: GECCO '19: Proceedings of the Genetic and Evolutionary Computation Conference Companion
• ISBN: 9781450367486
• DOI: https://doi.org/10.1145/3319619.3322054
• Public URL: http://researchrepository.napier.ac.uk/Output/2680883