Innovative automatic optimization method for ejectors in fuel cell vehicles based on a combined optimization strategy

Abstract

Ejectors exhibit significant advantages in the field of fuel cell vehicles, playing a crucial role in promoting their development. Their fixed structure results in non-parasitic power consumption, yet this also poses greater challenges for optimizing their structural parameters across different application scenarios. However, most of the research focuses on the positions such as the nozzle and mixing section, and there is no effective method to determine the values of all parameters. In order to solve the above problems, an innovative automatic optimization method using a combined optimization strategy (COS) with a weight factor is proposed for achieving multi-objective optimization of ejectors. The COS combines parametric modeling, computational fluid dynamics, approximate modeling techniques, and multi-objective optimization to tune the full parameters. The results indicate that the COS achieves a high performance prediction accuracy with an R2 value of 0.9711 and a root mean square error of 9.23E-6. Furthermore, in the case of 300 computational samples, the computational time is reduced by 54.7%. The entrainment ratio has been increased to 4.17 times its pre-optimization level. The novel method not only ensures the simulation accuracy but also significantly enhances computational efficiency, making it a powerful tool for guiding the production and optimization of ejectors.