Multi-fidelity model construction method, device and equipment

The invention provides a multi-fidelity model modeling method. The method comprises the following steps: determining a correlation function between a to-be-constructed multi-fidelity model and a constructed low-fidelity model; obtaining a low-fidelity sample set and a high-fidelity sample set; wherein the low-fidelity sample set has target sample points having the same abscissa as the sample points in the high-fidelity sample set; and calculating parameters of the correlation function according to the low-fidelity sample set and the high-fidelity sample set to obtain a prediction function of the multi-fidelity model, so that the multi-fidelity model performs prediction according to the prediction function. According to the multi-fidelity model modeling method provided by the invention, the model construction steps are simplified, and the multi-fidelity model can be quickly and accurately constructed. 本申请提供一种多保真度模型建模方法。所述方法包括：确定待构建的多保真度模型与已构建的低保真度模型之间的相关函数；获取低保真度样本集和高保真度样本集；其中，所述低保真度样本集中存在与所述高保真度样本集中各个样本点的横坐标相