Research on Sequence Sampling Method Based on Surrogate Model

The cooling fan is crucial for engine heat dissipation, and the blades, as an important component of the fan, have a significant impact on its performance. This paper focuses on the blade section, namely the cascade, as the research object. To address issues related to complex modeling and low accuracy in the study, a sequence sampling method based on local complexity and sample sparsity is proposed. This approach is contrasted with conventional sampling techniques employing standard test functions. The findings demonstrate that with fewer samples, the suggested approach can create a high-precision surrogate model. Lastly, real-world engineering applications are used to verify the efficacy of this approach.