Study on Method of Robust Multidisciplinary Collaborative Decision for Product Design

摘要

A robust method for decentralized multidisciplinary collaborative decision was presented which adopts interval to describe the uncertainty of decision variables and non-cooperative game theory to model the relationship between design teams under decentralized decision environment. Considering the Rational Reaction Sets (RRS) from non-cooperation game theory as design constraint in the decision space, multidisciplinary collaborative decision is described as interval based constraint solving problem and then a two-step solving approach was proposed to obtain final robust decision scheme of product design. The first step is to eliminate initial range of decision variables using consistency algorithm and the second step is to search the robust decision point in consistent interval with design capability indices as judgment rule of robustness. A new kind of feasibility censor based immune chaotic algorithm for model solving was designed. Design of bear and rotor system involves complex coupled relationship of dynamic and tribology and is a typical multidisciplinary conflict and decouple problem. A robust and powerful decision method between different disciplines can not only quicken design process of bear and rotor, but also improve the design quality. To show the problem, a typical elastic bear and rotor system with single disk is used as example to validate the effectiveness and reliability of proposed decision approach.