Multi agent system usually has to solve problems with some specifi c constraints, which are possibly to be over constraint. Moreover multiagent system has to utilize existing agents with their limited constraints knowledge. To address this problem, we use one distributed constraint satisfaction problem algorithm, asynchronous incremental relaxation in extreme kitchen domain. This algorithm uses multivalue backtrack with threshold value has proven able to solve over constraints problem by suggesting goal with the least constraint violation.
Copyrights © 2010