Gait optimization of AIBO robot based on interactive evolutionary computation

  title={Gait optimization of AIBO robot based on interactive evolutionary computation},
  author={John R. Eperjesi},
  journal={2008 6th International Symposium on Applied Machine Intelligence and Informatics},
  • J. Eperjesi
  • Published 7 March 2008
  • Computer Science, Biology
  • 2008 6th International Symposium on Applied Machine Intelligence and Informatics
This Master theses offers summary of methods used in interactive evolutionary computation. Their main aim is to reduce human fatigue during the evolution process. Some of these methods are then applied in evolution of AIBO robot gait pattern and compared. This work also includes summary of projects which used non-interactive methods for gait evolution. 

Figures from this paper

A Review of Gait Optimization Based on Evolutionary Computation
The techniques used in evolution-based gait optimization are reviewed, including why Evolutionary Computation techniques should be used, how fitness functions should be composed, and the selection of genetic operators and control parameters.
Distributed and Parallel Optimization of Captured Human Motions for the Generation of Robust and Stable Motions on Simulated Humanoid Robots
This thesis is about an extension of the existing framework to generate stable motions for simulated humanoid robots using a motion capture framework and then optimizing the motions efficiently in a distributed and parallel environment.


Evolutionary Gait-Optimization Using a Fitness Function Based on Proprioception
This paper presents a new approach to optimize gait parameter sets using evolutionary algorithms. It separates the crossover-step of the evolutionary algorithm into an interpolating step and an
Autonomous evolution of gaits with the Sony Quadruped Robot
This paper reports the results of autonomous evolution of dynamic gaits for the Sony Quadruped Robot, where fitness is determined using the robot's digital camera and infrared sensors.
Evolving the Walking Behaviour of a 12 DOF Quadruped Using a Distributed Neural Architecture
The main target of this paper is to show that the distributed architecture for the general control of robots can be applied to complex dynamic tasks like walking and show how a completely neural and distributed controller can be obtained for a robot ascomplex as Aibo on a task as complex as walking.
Observations on the Evolution of Neural Networks for the Control of a Simulated Quadruped Robot
This study used a genetic algorithm to evolve neural network controllers for the walking behavior of a quadrupedal robot. A 3D physics simulator (Breve by Jon Klein, using the ODE for physics
Evolving robot behavior via interactive evolutionary computation: from real-world to simulation
An interactive evolutionary system for behavior design is presented that allows a behavior engineer to interactiv ely teac h a robot obstacle avoidance behavior in minutes rather than in hours, thus reducing the amount of fatigue experienced by the behavior engineer.
Interactive evolutionary computation: fusion of the capabilities of EC optimization and human evaluation
This paper features a survey of about 250 IEC research papers and discusses the IEC from the point of the future research direction of computational intelligence.
ANDREJKOVÁ: Neurónové siete – inžiniersky prístup (Dopredné neurónové siete)
  • Elfa-Press,
  • 1996
Redukcia únavy používate a IEC pomocou neurónovej siete pri grafickom návrhu reklamných bannerov, Diplomová práca
  • Košice: Technická univerzita,
  • 2006
Autonomous evolution of SAMI 2008 • 6th International Symposium on Applied Machine Intelligence and Informatics 239 gaits with the Sony quadruped robot
  • Sony Corporation,
  • 1999
Using Artificial Neural Networks to Construct a Meta-Model for the Evolution of Gait Patterns
  • Universtity iof Dortmund,
  • 2002