Optimization method of discontinuous domain for finite element calculation

Abstract

The invention discloses an optimization method of a discontinuous domain for finite element calculation, is used for solving the problem that an optimal solution is difficult to find by a traditionalalgorithm due to large finite element calculation amount, and also provides a method for optimizing the discontinuous domain. According to the method, the method comprises the steps of: firstly, identifying an initial design space through a Latin hypercube sampling and K-means clustering algorithm, and generating k non-continuous feasible sub-design spaces for narrowing an optimization range, so that computing resources are saved for the optimization algorithm; and then optimizing the discontinuous sub-design space by using a multi-population particle swarm algorithm with a competition mechanism and mutation operation. Simulation results show that compared with other algorithms, the improved multi-population particle swarm optimization algorithm can improve the global search capability ofparticles and the utilization rate of computing resources, and the precision of optimization results is greatly improved.

Description

technical field [0001] The invention belongs to the technical research field of optimization algorithms, and in particular relates to an optimization method for discontinuous domains used in finite element calculations. Background technique [0002] As a powerful calculation method, the finite element method has been widely used in many engineering design fields, such as motor design, automobile, machinery manufacturing, shipbuilding and civil engineering. Optimization of structures by finite elements has resulted in improved product performance. However, in actual engineering, most of them are optimization problems with constraints, and finite element calculations often face the problem of excessive calculation, which makes it difficult for optimization algorithms to find the global optimal solution. [0003] In order to reduce the amount of calculation, Hu Chen proposed to identify the design space based on the K-means clustering algorithm and optimize the subspaces seque...

AI Technical Summary

Problems solved by technology

However, using the distance to determine the subspace will make the variables with a large value range cover the variables with a small range, and the optimization of equal resources in the subspace will consume too much computing resources in the poor subspace, and make the algorithm search accuracy of the excellent subspace reduce

[0004] In Engineering Optimization with Constrained Domains, Yongfei Xue(Y.Xue,Y.Wang and D.Shang, "Parameter Optimization of Hydrocracker using Multi-block Kriging Metamodelingwithin Discontinuous Operating Space," 2019 12th Asian Control Conference(ASCC), Kitakyushu -shi, Japan, 2019, pp.254-259.) In order to improve the accuracy of the kriging model, k subspaces satisfying the constraints are obtained from the constrained design space, and the subspaces are modeled separately and the resources are optimized. However, it will also cause a waste of computing resources

Images