Search

Abstract: This paper proposes a hybrid methodology that exploits strengths of wavelet analysis and support vector machine model in forecasting time series, and deals with the application of proposed methodology in manufacturing time series forecasting. This method is characteristic of the preprocessing of sample data using wavelet transformation for forecast, i.e., the data sequence of evolvement of share of some sectors in manufacturing is first mapped into several time-frequency domains, and then a support vector machine is established for each domain. The final forecasting results are the algebraic sums of all the forecasted components obtained by respective support vector machine models corresponding to different time-frequency domains. Nevertheless, one of disadvantages of the method is dilemma of selection of values of parameters in support vector machine because the way of selecting values for the parameters will affect the generalization performance remarkably. In this paper, chaos optimization is applied to accomplish selection of values of parameters. Results of experiments based on gross values of textile product in Japan suggest that this hybrid method can both achieve higher accuracy in manufacturing forecasting. [Copyright &y& Elsevier]

Fitting algorithms play an important role in the whole measuring cycle in order to derive a measurement result. They involve associating substitute geometry to a point cloud obtained by an instrument. This situation is more difficult in the case of non-linear geometry fitting since iterative method should be used. This article addresses this problem. Three geometries are selected as relevant case studies: circle, sphere and cylinder. This selection is based on their frequent use in real applications; for example, cylinder is a relevant geometry of an assembly feature such as pin-hole relationship, and spherical geometry is often found as reference geometry in high precision artifacts and mechanisms. In this article, the use of Chaos optimization (CO) to improve the initial solution to feed the iterative Levenberg–Marquardt (LM) algorithm to fit non-linear geometries is considered. A previous paper has shown the performance of this combination in improving the fitting of both complete and incomplete geometries. This article focuses on the comparison of the efficiency of different one-dimensional maps of CO. This study shows that, in general, logistic-map function provides the best solution compared to other types of one-dimensional functions. Finally, case studies on hemispheres and industrial cylinders fitting are presented. [ABSTRACT FROM AUTHOR]

Abstract: The aim of this paper is to propose and compare different one-dimensional maps as chaotic search patterns in the constraint nonlinear optimization problems. For this purpose, about 10 one-dimensional maps are introduced that can be used as search pattern in chaos optimization algorithms. We apply these maps in specific optimization algorithm (weighted gradient direction based chaos optimization algorithm) and compare them based on numerical simulation results. [Copyright &y& Elsevier]