Decomposed surrogate based optimization of carbon-fiber bicycle frames using Optimum Latin Hypercubes for constrained design spaces

摘要

Ideal bicycle frames need to be lightweight, laterally and torsionally stiff, withstand high impact forces and be reliable against fatigue failure. These characteristics can be measured by real life longevity tests. In the present work a bespoke optimization software framework for the design of optimal bicycle frames with respect to the EFBe standard taking into account rider specific demands is developed. For improved efficiency the size and shape optimization are decomposed. The optimization scheme is realized by means of surrogate based optimization where the design of experiments are Optimum Latin Hypercubes created in the feasible portion of the design space.