Neuroevolutive Algorithms for Learning Gaits in Legged Robots
Gait generation for legged robots is a challenging task typically requiring either a hand-tuning design or a kinematic model of the robot morphology to compute the movements, generating a high computational and time efforts. Neuroevolution algorithms with the ability to learn network topologies, suc...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2019-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/8852635/ |
| _version_ | 1818936338436587520 |
|---|---|
| author | Pablo Reyes Maria-Jose Escobar |
| author_facet | Pablo Reyes Maria-Jose Escobar |
| author_sort | Pablo Reyes |
| collection | DOAJ |
| description | Gait generation for legged robots is a challenging task typically requiring either a hand-tuning design or a kinematic model of the robot morphology to compute the movements, generating a high computational and time efforts. Neuroevolution algorithms with the ability to learn network topologies, such as Neuroevolution of Augmenting Topologies (NEAT), Hypercube-based NEAT (HyperNEAT), and τ-HyperNEAT, have been used in the computational community to learn gaits in legged robots. An extended version of HyperNEAT called ES-HyperNEAT, where the substrate hosting the nodes of the neural network evolves in shape, seems to be a promising algorithm to evaluate gait learning tasks. Using two four-legged robot platforms with different degrees of freedom (Quadratot and ARGOv2), we compared the performance of a variety of neuroevolution algorithms based on HyperNEAT for learning gaits. The comparative analysis of the results reveals that the three evaluated algorithms, HyperNEAT, τ-HyperNEAT and ES-HyperNEAT, successfully generate gaits given a fitness function. In particular, for the Quadratot platform, ES-HyperNEAT learns faster and better than the other two methods, a result that is not observed in the ArgoV2. Additionally, ES-HyperNEAT produces more phase changes between joints movement, resulting in more natural robot movements. Finally, ES-HyperNEAT produces complex substrates from simple Compositional Pattern Producing Networks, CPPN-NEAT, allowing the simplification of the underlying neural network. |
| first_indexed | 2024-12-20T05:34:29Z |
| format | Article |
| id | doaj.art-3713e2649b794b5c9835a5d0954fee60 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-20T05:34:29Z |
| publishDate | 2019-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-3713e2649b794b5c9835a5d0954fee602022-12-21T19:51:39ZengIEEEIEEE Access2169-35362019-01-01714240614242010.1109/ACCESS.2019.29445458852635Neuroevolutive Algorithms for Learning Gaits in Legged RobotsPablo Reyes0Maria-Jose Escobar1https://orcid.org/0000-0003-1563-624XDepartment of Electronic Engineering, Universidad Técnica Federico Santa María, Valparaíso, ChileDepartment of Electronic Engineering, Universidad Técnica Federico Santa María, Valparaíso, ChileGait generation for legged robots is a challenging task typically requiring either a hand-tuning design or a kinematic model of the robot morphology to compute the movements, generating a high computational and time efforts. Neuroevolution algorithms with the ability to learn network topologies, such as Neuroevolution of Augmenting Topologies (NEAT), Hypercube-based NEAT (HyperNEAT), and τ-HyperNEAT, have been used in the computational community to learn gaits in legged robots. An extended version of HyperNEAT called ES-HyperNEAT, where the substrate hosting the nodes of the neural network evolves in shape, seems to be a promising algorithm to evaluate gait learning tasks. Using two four-legged robot platforms with different degrees of freedom (Quadratot and ARGOv2), we compared the performance of a variety of neuroevolution algorithms based on HyperNEAT for learning gaits. The comparative analysis of the results reveals that the three evaluated algorithms, HyperNEAT, τ-HyperNEAT and ES-HyperNEAT, successfully generate gaits given a fitness function. In particular, for the Quadratot platform, ES-HyperNEAT learns faster and better than the other two methods, a result that is not observed in the ArgoV2. Additionally, ES-HyperNEAT produces more phase changes between joints movement, resulting in more natural robot movements. Finally, ES-HyperNEAT produces complex substrates from simple Compositional Pattern Producing Networks, CPPN-NEAT, allowing the simplification of the underlying neural network.https://ieeexplore.ieee.org/document/8852635/Legged robotsgait learningneuroevolutionNEATHyperNEATCPPN |
| spellingShingle | Pablo Reyes Maria-Jose Escobar Neuroevolutive Algorithms for Learning Gaits in Legged Robots IEEE Access Legged robots gait learning neuroevolution NEAT HyperNEAT CPPN |
| title | Neuroevolutive Algorithms for Learning Gaits in Legged Robots |
| title_full | Neuroevolutive Algorithms for Learning Gaits in Legged Robots |
| title_fullStr | Neuroevolutive Algorithms for Learning Gaits in Legged Robots |
| title_full_unstemmed | Neuroevolutive Algorithms for Learning Gaits in Legged Robots |
| title_short | Neuroevolutive Algorithms for Learning Gaits in Legged Robots |
| title_sort | neuroevolutive algorithms for learning gaits in legged robots |
| topic | Legged robots gait learning neuroevolution NEAT HyperNEAT CPPN |
| url | https://ieeexplore.ieee.org/document/8852635/ |
| work_keys_str_mv | AT pabloreyes neuroevolutivealgorithmsforlearninggaitsinleggedrobots AT mariajoseescobar neuroevolutivealgorithmsforlearninggaitsinleggedrobots |