International Journal of Industrial Mathematics
Volume:7 Issue: 4, Autumn 2015

The length of equal minimal and maximal blocks has e ected on logarithm-scale logarithm against sequential function on variance and bias of de-trended uctuation analysis, by using Quasi Monte Carlo(QMC) simulation and Cholesky decompositions, minimal block couple and maximal are founded which are minimum the summation of mean error square in Horest power.

The use of third-party logistics (3PL) providers is regarded as new strategy in logistics management. The relationships by considering 3PL are sometimes more complicated than any classical logistics supplier relationships. These relationships have taken into account as a well-known way to highlight organizations'' flexibilities to regard rapidly uncertain market conditions, follow core competencies, and provide long-term growth strategies. Choosing service providers has been considered as a notable research area in the last two decades. The review of the literature represents that neural networks have proposed better performance than traditional methods in this area. Therefore, in this paper, a new enhanced artificial intelligence (AI) approach is taken into consideration to assist the decision making for the logistics management which can be successfully presented in cosmetics industry for long-term prediction of the real performance data. The presented AI approach is based on modern hybrid neural networks to improve the decision making for the 3PL selection. The model can predict the overall performance of the 3PL according to least squares support vector machine and cross validation technique. In addition, the proposed AI approach is given for the 3PL selection in a real case study for the cosmetics industry. The computational results indicate that the proposed AI approach provides high performance and accuracy through the real-life situations prediction along with comparing two other two well-known AI methods.

The determination of an unknown boundary condition, in a nonlinaer inverse diffusion problem is considered. For solving these ill-posed inverse problems, Galerkin method based on Sinc basis functions for space and time will be used. To solve the system of linear equation, a noise is imposed and Tikhonove regularization is applied. By using a sensor located at a point in the domain of $x$, say $x=a''$, and determining $u(a'',t)$ a stable solution will be achived. An illustrative example is provided to show the ability and the efficiency of this numerical approach.

Singular perturbation problems have been studied by many mathematicians. Since the approximate solutions of these problems are as the sum of internal solution (boundary layer area) and external ones, the formation or non-formation of boundary layers should be specified. This paper, investigates this issue for a singular perturbation problem including a first order differential equation with general non-local boundary condition. It needs to say that it is simple for local boundary conditions and there is no difficulty. However, the formation of boundary layers for non-local case is not as stright forward as local case. To tackle this problem generalized solution of differential equation and some necessary conditions are used.

It is well known fact that binary relations are generalized mathematical functions. Contrary to functions from domain to range, binary relations may assign to each element of domain two or more elements of range. Some basic operations on functions such as the inverse and composition are applicable to binary relations as well. Depending on the domain or range or both are fuzzy value fuzzy set, interval fuzzy value fuzzy set or fuzzy number value fuzzy set, define of the fuzzy relation is different. Given a fuzzy relation, its domain and range are fuzzy number value fuzzy sets. In this paper, initially we define fuzzy number value fuzzy sets and then propose fuzzy number-valued fuzzy relation (FN-VFR). We also introduce property of reflexive, symmetric, transitive and equivalence relation of FN-VFR. As follow, we prove some theorems for FN-VFR with property of reflexive, symmetric and transitive. Also, we show examples for FN- VFR.

This article is devoted to study of the autoconvolution equations and generalized Mittag-Leffler functions. These types of equations are given in terms of the Laplace transform convolution of a function with itself. We state new classes of the autoconvolution equations of the first kind and show that the generalized Mittag-Leffler functions are solutions of these types of equations. In view of the inverse Laplace transform, we use the Schouten-Vanderpol theorem to establish an autoconvolution equation for the generalized Mittag-Leffler functions in terms of the Laplace and Mellin transforms. Also, in special cases we reduce the solutions of the introduced autoconvolution equations with respect to the Volterra $\mu$-functions. Moreover, more new autoconvolution equations are shown using the Laplace transforms of generalized Mittag-Leffler functions. Finally, as an application of the autoconvolution equations in thermodynamic systems, we apply the Laplace transform for solving the Boltzmann equation and show its solution in terms of generalized Mittag-Leffler functions.

The paper presents the semi-numerical solution for the magnetohydrodynamic (MHD) viscous flow due to a stretching sheet caused by boundary layer of an incompressible viscous flow. The governing partial differential equations of momentum equations are reduced into a nonlinear ordinary differential equation (NODE) by using a classical similarity transformation along with appropriate boundary conditions. Both nonlinearity and infinite interval demand novel the mathematical tools for their analysis. The solution of the resulting third order nonlinear boundary value problem with an infinite interval is obtained using fast converging Dirichlet series method and approximate analytical method viz. method of stretching of variables. These methods have the advantages over pure numerical methods for obtaining the derived quantities accurately for various values of the parameters involved at a stretch and they are valid in much larger parameter domain as compared with HAM, HPM, ADM and the classical numerical schemes. Also, these methods require less computer memory space as compared with pure numerical methods.

The objective of this paper is applying the well-known exact operational matrices (EOMs) idea for solving the Emden-Fowler equations, illustrating the superiority of EOMs over ordinary operational matrices (OOMs). Up to now, a few studies have been conducted on EOMs; but the solved differential equations did not have high-degree nonlinearity and the reported results could not strongly show the excellence of this new method. So, we chose Emden-Fowler type differential equations and solved them utilizing this method. To confirm the accuracy of the new method and to show the preeminence of EOMs over OOMs, the norm 1 of the residual and error function for both methods are evaluated for multiple $m$ values, where $m$ is the degree of the Bernstein polynomials. We report the results by some plots to illustrate the error convergence of both methods to zero and also to show the primacy of the new method versus OOMs. The obtained results demonstrate the increased accuracy of the new method.