| |
Exploring the T-Maze: Evolving Learning-Like Robot Behaviors using CTRNNs
|
|
3 Experiment 1: Simple T-Maze
In the first experiment a robot has to navigate a simple T-maze (fig. 2). The
experiment is carried out in a realistic simulation of the Khepera robot (fig.
1(a)) based on sensor sampling [6] and adding 5% uniform noise to the sampled
values. Initially the robot is positioned as shown in figure 2, and the task is to
find and stay on the black reward-zone which can be positioned in either the left
or the right arm of the maze. The position of the reward-zone stays fixed during
each epoch. The robot is tested for 4 epochs of 5 trials each - two epochs with
the reward-zone in each arm of the maze. The neural network controlling the
robot is initialized (by setting the state of each neuron to zero) at the beginning
of each epoch but not between trials within the same epoch. This means that
the robot can potentially build up and store information in the dynamic state
of the network between trials within the same epoch. The optimal behavior of
the robot in this environment is to use the first trial of each epoch to locate
and “remember” the position of the reward-zone, and thereafter move directly
towards it for the remaining trials of the epoch. To put additional evolutionary
pressure on this behavior, the number of available sensory-motor steps is 360
in the first trial of each epoch and only 180 in the remaining 4 trials. Given
the size of the maze this means that the robot only has time to explore the
whole maze during the first trial of each epoch. In addition a poison-zone (white
square in figure 4(b)) is positioned opposite of the reward-zone in the last 4
but not the first trial of each epoch. An individual is immediately killed if it
steps over the poison-zone. The fitness function is simply the sum of trials an
individual ends its life inside the reward-zone. Since each individual is tested for
a total of 20 trials the maximal possible fitness is 20. Notice that there is no
direct pressure on evolving fast moving robots given this fitness function. This
is however compensated by the fact that the number of steps in each trial is
limited and has been adjusted to fit the size of the environment.
Page 1, 2, 3, 4, 5, 6
Tech Materials (Free)
Robot Behaviors Exploring the T-Maze: Evolving Learning-Like Robot Behaviors using CTRNNs
Humanoid Robotics A Biochemical Subsystem for a Humanoid Robot
Industrial Automation Systems Applying Agents for Engineering of Industrial Automation Systems
Robot Team Cooperation A Descriptive Model of Robot Team and the Dynamic Evolution of Robot Team Cooperation
Kuka Robots For ONU ONU Robotics Technology Center of Excellence, powered by KUKA Robotics Corporation
Augmented reality Annotation System for Robotic Application
Modular Robots Self-Reconfiguration Planning Of Identical Modules
Autonomous robots A New Approach To Robotics
Robotic Mounting Flat Panel Displays With Robotic Mounting
Calibration of Industrial Robots A Photogrammetric Robot Calibration System Based On Off-The-Shelf Low Cost Hardware Components
More... Amazon Books
|
|
|
|
