Abstract: In order to realize the optimization selection and utilization of manufacturing resource in the development process of complicated product, a kind of multipurpose optimization decision-making method based on development process was brought forward. Facing towards hierarchy and decomposition of task in the development process, the restrictions and evaluation criterions of resource selection were combined with product development task, and the practicability of resource selection was ensured. The establishment of product development environment and process realization of resource dispatch based on agent made preparation for resource sharing and optimization utilization. By improved method of multipurpose optimization decision-making, combining with two utility evaluation functions under the standard of gain target and profit loss target, the principle and process of resource optimization selection were depicted in detail which based on the targets such as response time, service quality, cost and security, etc, and the sort and selection of resource dispatch were realized. Finally, the feasibility of algorithm was verified by a real instance in engineering field.

Key words: complicated product, manufacturing resource, multipurpose decision-making

摘要： 为实现复杂产品开发过程中制造资源的优化选择和使用，提出一种基于分层的多目标优化 决策方法。面向过程的任务分层分解技术，将影响资源选择的约束因素和评价标准与特定的产品开 发任务相结合，保证所选择资源的实用性。并行协同产品开发环境的建立以及基于Agent技术的资源 调度过程实现，为制造资源共享和优化使用准备了条件。通过改进的多目标优化决策方法结合增益 目标、损益目标准则下两个效用评价函数，详细地描述了服务响应时间、服务质量、成本、安全性 等约束目标下资源优化选择的原理和过程，实现了资源调度的“排序选优”。结合工程实例，验证 了该算法的可行性。

关键词: 复杂产品, 制造资源, 多目标决策