Sparse Hybrid Uncertain Variable Optimization Method and Application

Peng ZHANG,Xiao-jian LIU,Shu-you ZHANG,Le-miao QIU,Guo-dong YI
DOI: https://doi.org/10.3785/j.issn.1008-973x.2019.03.004
2019-01-01
Abstract:Sparse hybrid uncertainties in the distribution of random variables brings the problem of hard border detection and calculation results distortion in complex product design. A sparse hybrid uncertain variable optimization method based on Chebyshev approximation was proposed. Firstly, the maximum likelihood estimation was utilized to construct the probability density function of a sparse hybrid uncertain variable under a given distribution, and its distribution parameter corresponding to the given distribution was preliminarily determined. Secondly, based on the Bayesian information criterion, the information loss of the corresponding distribution was calculated, and the most suitable distribution and parameters of the sparse mixed uncertain variables were further determined. Thirdly, in order to solve the computational distortion caused by interval expansion of the traditional interval analysis method, Chebyshev approximation was used to optimize the objective function and the improved HL-RF algorithm was used to obtain the reliability index and the failure probability value; while meeting the design requirements, the design goal of light weight and rigidity retention was effectively realized. Finally, the effectiveness of the proposed method is verified by numerical examples and the optimization design of a high-speed press slider.
What problem does this paper attempt to address?