作業割り当て問題に対する自律分散型スケジューリングシステム

抄録

This paper deals with assigning N operations on parallel H machines. The earliest start time, the latest finish time and the processing time are defined for each operation. The problem is to minimize the deviation of each processing time from the desirable time interval. We propose two autonomous decentralized scheduling algorithms to solve this problem. In these algorithms, an operation is transferred from one machine to another machine on the basis of information exchanges and cooperation among subsystems. By examining numerical results, a comparison is made between two algorithms from the viewpoints of accuracy and computation time. Consequently, abetter algorithm is found to be the one based on negotiations of one subsystem with the others.