A Multi-Domain Physical Model Design Optimization Method Considering Uncertainties

Large-scale To reduce the impact of uncertainty on the design results, A Multi-Domain Physical Model Design Optimization Method Considering Uncertainties is proposed. The method use XML to incorporate uncertainty into the design variables and design parameters of the Modelica model. The performance objective function and state function are obtained by solving the sequence of performance variables. A hybrid agent model based on incremental sequence sampling is used to reduce calculation of optimization. Finally, An engineering application test are demonstrated to verify the effectiveness of the proposed method.