A Hybrid GP/GA Approach for Co-Evolving Controllers and Robot Bodies to Achieve Fitness-Specified Tasks.
Wei-Po LeeJohn HallamHenrik Hautop LundPublished in: International Conference on Evolutionary Computation (1996)
Keyphrases
- fitness function
- genetic programming
- genetic algorithm
- genetic algorithm ga
- evolutionary algorithm
- evolutionary process
- evolutionary computation
- robotic systems
- evolutionary robotics
- mobile robot
- service robots
- autonomous robots
- evolution process
- neural network
- crossover and mutation
- human operators
- fitness landscape
- robot navigation
- real robot
- manipulation tasks
- parallel genetic algorithm
- vision system
- reinforcement learning
- real time
- human robot interaction
- path planning
- navigation tasks
- legged robots
- real world environments
- wheeled mobile robot
- controller design
- population size
- robot control
- genetic operators
- humanoid robot
- particle swarm optimization pso