Unconventional optimization for achieving well-informed design solutions for the automobile industry

ABSTRACT

In practice, engineering optimization problems come in various complexities that are different from the usual problem formulations used in a single academic study. To solve such unconventional description of problems, optimization experts must use different tools and tricks first to reformulate the problems as structured optimization problems and then to use an appropriate adaptation of existing optimization algorithms to solve them. In this article, a design optimization case study from the automobile industry is narrated, stating three different problem solving scenarios arising sequentially after each problem is solved. The article shows how each problem description can be formulated as a suitable optimization problem and how a suitable and existing optimization algorithm can be modified to find the respective solutions of interest to the engineer. The sequential reformulation of problems and their solutions not only allow engineers eventually to discover the final solutions of interest, the whole process is extremely rewarding to them in terms of knowledge gathered about alternative potential solutions of the problem. Although applied to a single automotive design problem in this article, the philosophy and methods described here are generic and should motivate both engineering optimization researchers and the industry to pursue more such studies.

KEYWORDS:

• Unconventional optimization
• optimal design
• industrial design
• multi-objective optimization
• constraint relaxation

Acknowledgment

The views presented in this article do not necessarily reflect those of the sponsor, General Motors, whose support is gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research was supported by General Motors Corporation.