1 / 13

Evolutionary Robotics

Evolutionary Robotics. First attempt to evolve robot morphology and control. Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. Artificial Life IV , 28-39. Evolutionary Robotics. First attempt to evolve robot morphology and control. Evolution performed on

rcrowe
Download Presentation

Evolutionary Robotics

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Evolutionary Robotics First attempt to evolve robot morphology and control Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. ArtificialLife IV, 28-39.

  2. Evolutionary Robotics First attempt to evolve robot morphology and control Evolution performed on Connection Machine-5: 32 processors connected in hypercube topology. Population size = 300 # of generations = 100 ~ 3 hours for a typical run Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. ArtificialLife IV, 28-39.

  3. Evolutionary Robotics First attempt to evolve robot morphology and control Genotypes: Phenotypes: Genotype-to-phenotype mapping: Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. Artificial Life IV, 28-39.

  4. Evolutionary Robotics First attempt to evolve robot morphology and control Genotype: Node labels: body part dimensions joint-type: hinge, ball/socket… joint-limits recursive-limit (RL): if ( RL>0 ): RL = RL-1 follow all outgoing edges. Edge labels: change in position change in orientation change in scale reflection terminal-only flag (TO): if ( TO==true && parent->RL==0 ): follow edge;

  5. Evolutionary Robotics First attempt to evolve robot morphology and control 2 3 1 Genotype: Node labels: body part dimensions joint-type: hinge, ball/socket… joint-limits recursive-limit (RL): if ( RL>0 ): RL = RL-1 follow all outgoing edges. Edge labels: position orientation scale reflection terminal-only flag (TO): if ( TO==true && parent->RL==0 ): follow edge; 5 6 4 8 9 7 11 12 10

  6. Evolutionary Robotics First attempt to evolve robot morphology and control Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. ArtificialLife IV, 28-39.

  7. Evolutionary Robotics First attempt to evolve robot morphology and control Non-local “Brain” Genotypes: Nested, directed, multigraphs: Each graph node contains another graph. Embedded graph describes the local neural circuitry for that body part. JX = joint sensor CX = contact sensor PX = photosensor EX = effector (X=degree-of-freedom)

  8. Evolutionary Robotics First attempt to evolve robot morphology and control Non-local “Brain” Neurons: Sum Product Divide Sum-threshold Greater-than Sign-of, Min, Max, Abs, If, Interpolate (itp) Sin, cos, atan, Log, Expt, Sigmoid, Integrate, Differentiate, Smooth, Memory Oscillate-wave (wav) Oscillate-saw (saw).

  9. Evolutionary Robotics First attempt to evolve robot morphology and control Genotypes: Nested, directed, multigraphs: Each graph node contains another graph. Embedded graph describes the local neural circuitry for that body part. JX = joint sensor CX = contact sensor PX = photosensor EX = effector (X=degree-of-freedom) Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. ArtificialLife IV, 28-39.

  10. Evolutionary Robotics First attempt to evolve robot morphology and control Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. ArtificialLife IV, 28-39.

  11. Evolutionary Robotics First attempt to evolve robot morphology and control Two robots compete: Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. ArtificialLife IV, 28-39.

  12. First attempt to evolve robot morphology and control Robots can compete within a single species (a-d), or robots can compete between species (e-g) What is the disadvantage of ‘all-vs-all’, ‘random’, and ‘tournament’? Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. ArtificialLife IV, 28-39.

  13. First attempt to evolve robot morphology and control Results from four evolutionary runs. Grey line: Average fitness of species 1; Black line: Average fitness of species 2. Sims, K. (1994) Evolving 3D Morphology And Behavior by Competition. ArtificialLife IV, 28-39.

More Related