CFM 2019

Multi-criteria optimization of process planning and part design
Séverine Durieux  1@  , Laurent Delolme  1@  , Charles Fortunet  1@  , Hélène Chanal  1@  , Emmanuel Duc  1, *@  
1 : Université Clermont Auvergne, SIGMA Clermont, Institut Pascal  (UCA, CNRS, SIGMA)
CNRS : UMR6602, ubp, SIGMA Clermont
* : Auteur correspondant

Currently, the product design and manufacturing processes is optimized jointly to meet the increasing level of industrialization performance and product service performance. These increases in performance translate into the expression of various often conflicting indicators related to each activity of the process and the service. The usual design and manufacturing process is a rather linear process, capable of optimizing each step individually and inherently, but it can not take into account a multi-criteria global optimization. Generally only one solution is proposed and optimized gradually throughout the process, the optimized design conditioning the optimization of the manufacturing.

We propose a multicriteria decision-making process based on the AHP and Topsys methods. As a first step, a genetic algorithm is used to produce a wide variety of solutions (design and manufacturing) from a parameterization of the studied part. For each solution, the indicators are evaluated and are aggregated in a macro-criterion according to an AHP approach, to model the behavior of the decision-maker. The user can finalize the choice from a sample of some solutions recognized as performing, using the Topsis method.

The approach is applied firstly to the process planning and secondly to the Design For Manufacturing of aeronautical parts.


Personnes connectées : 63