Some Experimental Results Relevant to the Optimization of Configuration and Planning Problems

Abstract

This communication deals with mass customization and the association of the product configuration task with the planning of its production process while trying to minimize cost and cycle time. We consider a two steps approach that first permit to interactively (with the customer) achieve a first product configuration and first process plan (thanks to non-negotiable requirements) and then optimize both of them (with remaining negotiable requirements). This communication concerns the second optimization step. Our goal is to evaluate a recent evolutionary algorithm (EA). As both problems are considered as constraints satisfaction problems, the optimization problem is constrained. Therefore the considered EA was selected and adapted to fit the problem. The experimentations will compare the EA with a conventional branch and bound according to the problem size and the density of constraints. The hypervolume metric is used for comparison.

Optimizing concurrent configuration and planning: A proposition to reduce computation time

“…In [7] and [8], we have described how user requirements (for example, number of seats belongs to [6,8] or due date is prior to 31/10/2010) can be processed while leaving some choices undecided (for example flying speed and range remain not set). These undecided characteristics are frequently set by commercial configuration systems with default values.…”

“…When the problem gets larger many authors, see [9] or [10], propose to use evolutionary algorithms to handle the problem. In this idea, we have proposed and discussed in [7] and [8] Algorithm. However, for larger problem (> 10 12 solutions) Pareto computing time is quantified in days.…”

“…To avoid the generation of unfeasible individuals, the domain of every remaining gene is updated by constraint filtering. Thanks to consult [7] and [8] for more details.…”

“…An average constraint level, that rejects 93% of solutions, allows 7.3*10 16 feasible solutions. Results of experimentations with other model sizes and other constraint levels can be consulted in [8]. Figure 5 shows the Pareto Fronts obtained with CFB-EA after 3 and 24 hours of computation.…”