Application method of complex shape-based self-adaptive proxy model in optimal design of battery modules

Abstract

The invention discloses a complex shape-based self-adaptive agent model optimization method and its application in battery module optimization design. Since traditional optimization methods are costly, long-term, and complex to solve black-box problems or engineering problems with complex performance functions, an adaptive proxy model optimization method based on complex shapes is proposed and applied to the optimization of battery modules. designing. Firstly, the ANSYS geometric optimization model of the air-cooled battery module is constructed; secondly, the CFD calculation is carried out according to the ANSYS geometric optimization model; then, the surrogate model is established by using the experimental design method and the approximation technology; finally, the air-cooled battery module optimize. The invention applies the complex shape method to the iteration of adding points in the sequence of the proxy model, and has the advantages of less added samples, faster convergence speed, high local precision and high optimization efficiency than the traditional optimization method of adding points in the sequence, and has broad application prospects.

Description

technical field [0001] The invention belongs to the technical field of engineering optimization design, and in particular relates to an application method of a complex shape-based self-adaptive agent model in battery module optimization design. Background technique [0002] Black-box problems and optimization problems with complex performance functions in the engineering field are usually difficult to optimize directly. Surrogate-Based Optimization (SBO) has been used to solve implicit functions and complex and time-consuming expensive model optimization problems, such as aerospace systems involving aerodynamics, structural optimization, turbine blades, etc. Proven to be a proven method. The surrogate model was proposed by Kleijnen in 1975. Sacks et al. realized the stochastic process of deterministic output based on computer experiments, so that it is more convenient to predict. This work is considered to be the beginning of approximate optimization research. After years ...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is: in order to solve the problem that the traditional proxy model-based optimization method has a complicated process, a large number of adaptive points, and a high optimization cost, the present invention proposes a compound-based self-adaptive proxy model in the battery module Applied methods in optimal design

Images