Aircraft system optimization method based on dynamic multi-model fusion

Abstract

The invention discloses an aircraft system optimization method based on dynamic multi-model fusion, and belongs to the technical field of multidisciplinary optimization in aircraft design. The methodincludes the steps: fusing high-and-low-accuracy analysis models in aircraft system design by a Co-Kriging method, and replacing an original model to perform optimization design; updating and managinga Co-Kriging surrogate model by a double sampling method in the optimization process; improving local convergence ability of optimizing search through trust region sampling, improving global exploration ability of an optimizing process through predictive variance sampling, and guiding the optimizing process to rapidly converge an optimal solution. The method can overcome the shortcomings that a traditional optimization method can only process single-accuracy analysis model and is high in optimization cost, multi-source response information cannot be sufficiently utilized and the like, simulation models with different accuracies in aircraft system design can be efficiently fused and optimized, and the aircraft system optimization method is of great significance for reduction of aircraft system design optimization cost, improvement of system design performance and the like.

Description

technical field [0001] The invention relates to an efficient optimization design method for a complex aircraft system based on dynamic multi-model fusion, and belongs to the technical field of multidisciplinary optimization in aircraft design. Background technique [0002] With the development of computer technology, high-precision simulation models have been widely used in aircraft system design, such as finite element analysis FEA model, computational fluid dynamics CFD model, etc. While the high-precision simulation model effectively improves the analysis accuracy and design reliability, it also significantly increases the calculation cost. For example, it often takes several hours or even dozens of hours to complete an aerodynamic simulation analysis using a CFD model. Since traditional optimization methods (such as genetic algorithms, etc.) often need to call thousands of analysis models to realize the exploration of the design space, direct use of high-precision simula...

AI Technical Summary

Problems solved by technology

[0016] Aiming at the traditional surrogate model optimization method that can only optimize a single-precision analysis model, resulting in the high cost of aircraft system optimization and the inability to make full use of multi-source response information, the invention discloses an aircraft system optimization method based on dynamic multi-model fusion to solve The technical problem is: realize the multi-model fusion aircraft system optimization, improve the global convergence of the optimization process and reduce the calculation cost, and shorten the design cycle

Images