Evolving Humanoid Behaviors for Language Games

Evolutionary techniques are applied to develop the neural control of humanoid robots. These robots were designed to act as agents in embodied language games. The basic ingredients needed to bring forth the desired behaviors are described: an appropriate physical simulator of the robots, an interactive evolution environment and various analysis tools. A modular approach to neural control is taken and is supported by a corresponding evolutionary algorithm, such that complete neural control networks are composed of specific functional units, the so called neuro-modules. Examples of such modules are described and their use is demonstrated by means of two developed networks for a walking and a gesture behavior.

[1]  Robert J. Peterka,et al.  Comparison of human and humanoid robot control of upright stance , 2009, Journal of Physiology-Paris.

[2]  Frank Pasemann,et al.  SO(2)-Networks as Neural Oscillators , 2003, IWANN.

[3]  B. Yegnanarayana,et al.  Artificial Neural Networks , 2004 .

[4]  Luc Steels,et al.  Language Grounding in Robots , 2012, Springer US.

[5]  Stefano Nolfi,et al.  Evolutionary robotics , 1998, Lecture Notes in Computer Science.

[6]  Manfred Hild,et al.  Myon, a New Humanoid , 2012, Language Grounding in Robots.

[7]  Frank Pasemann,et al.  NERD Neurodynamics and Evolutionary Robotics Development Kit , 2010, SIMPAR.

[8]  Karl Johan Åström,et al.  PID Controllers: Theory, Design, and Tuning , 1995 .

[9]  Marco Y C Pang,et al.  Split-Belt Treadmill Stepping in Infants Suggests Autonomous Pattern Generators for the Left and Right Leg in Humans , 2005, The Journal of Neuroscience.

[10]  Manfred Hild,et al.  Myon: Concepts and Design of a Modular Humanoid Robot Which Can Be Reassembled During Runtime , 2011 .

[11]  Ansgar Büschges,et al.  Deriving neural network controllers from neuro-biological data: implementation of a single-leg stick insect controller , 2011, Biological Cybernetics.

[12]  Hiroaki Kitano,et al.  RoboCup: The Robot World Cup Initiative , 1997, AGENTS '97.

[13]  James A. Yorke,et al.  Dynamics: Numerical Explorations , 1994 .

[14]  Frank Pasemann,et al.  Neural control of a modular multi-legged walking machine: Simulation and hardware , 2012, Robotics Auton. Syst..

[15]  Frank Pasemann,et al.  Search Space Restriction of Neuro-evolution through Constrained Modularization of Neural Networks , 2010, ANNIIP.

[16]  Frank Pasemann,et al.  Dynamical Neural Schmitt Trigger for Robot Control , 2002, ICANN.

[17]  Frank Pasemann,et al.  Reflex-oscillations in evolved single leg neurocontrollers for walking machines , 2007, Natural Computing.

[18]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .