Rescheduling method based on distributed cooperative problem solving model

This paper proposes a rescheduling method based on a distributed cooperative problem solving model to cope with contingencies. Using several Planners with different search strategies, the rescheduling system tries to obtain a solution. In case the solution which satisfies all the constraints cannot be found, the proposed system detects some of the most promising positions where it is effective to apply relaxation of constraints, and displays candidate conditions to be relaxed to the human operator for the purpose of supporting his/her decision. This idea is applied to a train rescheduling problem, and the result of the experiment is shown.