Abstract: To extend the decoupling ability of traditional workflow engine, and to eliminate the data flow decoupling in multidisciplinary design process, a multi-objective collaborative data flow decoupling strategy was proposed, and a data flow decoupling ware was developed. In order to improve reusability of integrated design software and disciplinary model, from the perspective of collaborative design method framework, the modeling process was modified. By remaining unchanged of various sub-disciplinary models, design results of various disciplines were analyzed by using clustering method. Functions to judge multi-objective satisfaction rate were added in order to acquire reasonable design results.The application of small solid rocket design proved that the extended workflow engine could realize the process decoupling control of multidisciplinary design and obtain the global feasible solutions of design problem, thus the management disadvantages of traditional workflow method on multidisciplinary coupling design process were remedied.

Key words: multi-objective collaborative strategy, multidisciplinary design, process management, data flow, decoupling ware, small solid rocket

摘要： 为扩展传统工作流引擎的解耦能力，消除多学科设计过程中的数据流解耦，提出基于多目标协同的数据流解耦策略，并设计了数据流解耦器。为了提高集成设计软件和学科模型的可重用性，该方法从协同设计方法框架出发，对其建模过程进行了改进，在不改变各子学科设计模型的基础上，通过应用聚类方法对各学科设计结果进行分析，并增加了多目标满意度判断函数，以期获得合理的设计结果。通过在小型固体火箭总体设计中的应用，表明扩展后的数据流引擎能够实现多学科设计问题的过程解耦控制，并能够获得设计问题的全局可行较优解，从而弥补了传统工作流方法对多学科耦合设计过程管理的不足。

关键词: 多目标协同策略, 多学科设计, 过程管理, 数据流, 解耦器, 小型固体火箭