Multitarget Hierarchical Negotiation for Distributed Design in Collaborative Product Development

Abstract

Distributed design in collaborative product development involves interorganizational teams of different disciplines, often geographically separated, that work jointly for the interest of an overall goal. The teams must agree on the product design without fully disclosing individual design information. Price-based negotiation is an effective design methodology under such circumstances. This work improves the negotiation mechanism based on the price schedules decomposition algorithm (PSDA), which only allows bi-level negotiation with a single design target, in two regards. The augmented price schedules decomposition algorithm (APSDA) is proposed to simultaneously consider multiple design targets in negotiation. Decision models and negotiation protocols are developed for complex design tasks decomposed in a hierarchy. The resultant APSDA-based negotiation mechanism enables multi-target hierarchical distributed design. A test scenario implemented with multiagent technologies validates the effectiveness of this mechanism. This study solves automating engineering collaborations with secure information sharing.Note to Practitioners-In collaborative product development, designers often need to collaborate on an overall design goal with individual objectives and potential conflicts. Individual design rationales may not be completely transparent to each other. This work proposes an automatic negotiation mechanism to conduct distributed design in such a circumstance. This mechanism is implemented in a multiagent environment and tested with a scenario of distributed weight allocation in the development of a training simulator. The test results validate the effectiveness of the mechanism.

Dynamic Coalition Formation Based on Multi-sided Negotiation

“…On the other hand, many optimal models are proposed aiming at special application regions. The Tribase model [7] uses trust degree, acquaintance degree to form and adjust the coalition. Tong Xiang-Rong [8] presents a long term coalition maintenance algorithm based on fuzzy ally relation.…”