Technical Paper
Multidisciplinary Design Method for Off-Road Vehicle using Bayesian Active Learning
2024-04-09
2024-01-2595
In developing off-road vehicles, it is essential to create high drivability that can run on all ground surfaces and passenger comfort. In addition, durability performance is also indispensable, and a development method that considers a high-level combination of these performances is required. This paper proposes a method to search for the region where each performance is established in a complex domain combination problem. The proposed method is helpful for the initial stage design when the details of the specifications of the target vehicle are uncertain because it is possible to consider not only the specifications of the target vehicle but also the operation method, such as the selection of the driving road profile and the driving speed as design variables. The previous method has the advantage of efficient concurrent study to search feasible regions.