A proposal for decentralized cooperative decision-making in chemical batch operation

Abstract

The objects of this study are to discuss an intelligent solution of dynamic unit-assignment problems in the decentralized cooperative operation for a new type of chemical batch plant, i.e. a pipeless batch plant, and to propose a decentralized cooperative method of multiple-criteria decision-making in process planning. For these purposes, two techniques of Distributed Artificial Intelligence (DAI) and the Analytic Hierarchy Process (AHP) are applied. Negotiation is realized by using message passing based on the contract net model, and a method of decentralized dynamic unit-assignment makes local scheduling possible. By integrating such multiple local schedules, cooperative operation and management of the whole plant may be achieved. The AHP is applied to the decision-making in selecting a suitable unit among alternative ones. Multiple functional units, such as process managers, reactors and stations, make game-like decisions based on multiple criteria of static evaluation and dynamic evaluation. Such a system is proposed to realize the flexibility and adaptability required for future chemical batch plants, rather than optimality.