Obstacle avoidance of redundant manipulators using genetic algorithms
Ahmad Reza Khoogar, Joey K. Parker · 2002
The authors introduce a novel method for obstacle avoidance with redundant robot manipulators based on genetic algorithms (GAs), GAs are search procedures which duplicate many of the processes of natural genetics. GAs work with a coding of the parameters. The use of a 30-b ternary number to represent the moves associated with a three-degree-of-freedom robot is discussed. A finite number of incremental moves are used to guide the end-effector to the goal point through a collision-free path. Two example runs are described, and advantages and disadvantages of the method are discussed.>