A resource allocation problem on timed marked graphs: a decomposition approach

In complex man-made environments discrete event dynamic systems are frequently encountered, and a timed marked graph is widely accepted as a convenient tool to describe systems of this kind. We consider trade-offs of cost and performance on such a system. First we formulate an optimization problem and transform it into a mixed integer linear programming problem. To improve computational efficiency, we decompose the problem into two phases. In phase one we determine the optimal number of each resource to be adopted in the system, and in phase two we optimize the distribution of these resources over the system. Phase one is solved very quickly and approximately by the dominance relaxation through a binary search procedure. This also gives the estimate of error bounds. An illustrative example shows an application to a jobshop optimization problem, and numerical experiments are carried out for some sample problems.     

Hexagonal β-SiC Particles

Among small β-SiC particles (0.2 ∼ 0.4 μm in size) produced by chemical vapor reaction, hexagonal particles having diameters around 0.2 μm are sometimes found. When observed by TEM, they have image contrasts reminiscent of "multiply twinned particles". Analysis of diffraction patterns and image contrasts observed indicates that these particles consist of multiple twins with twin planes parallel to the basal plane. The characteristics of growth processes of such particles are discussed.     

An improvement of variable sampling plans based on Kullback-Leibler information

A new and convenient design procedure for single sampling plans by variables is given by Arizono and Ohta (1986) based on the Kullback-Leibler information. In this paper an improvement of the original design procedure is presented and the revised procedure is compared with the bayesian procedure.     

An information-theoretic approach to classifying sensory inspectors

Tests of product quality that depend exclusively on one or more of the five senses of a human inspector are known as sensory tests. Since the ability to make correct sensory judgments can vary substantially from person to person, it is important in the selection and training of inspectors to be able to distinguish between inspectors of differing sensory abilities. This paper proposes a new approach based on the Kullback-Leibler data for classifying inspectors by panel tests. A method of deciding the same size in the test inspection lot with given risks is also suggested. The recommended procedure is simple, practical and relatively effective.     

Determination of single-sampling-attribute plans based on membership functions

Single-sampling-attribute plans which make the producer's risk less than or equal to α (that is,  α) and the consumer's risk less than or equal to β (that is, β) are called exact sampling plans. In using exact sampling plans, however, cases often arise in which large sample size is needlessly sought. A minimax criterion which relaxes the conditions on the risks for reducing the sample size was. introduced by Fujino and Okuno. This paper presents a design procedure for relaxing the conditions on the risks by the membership functions used in fuzzy methodologies. The proposed procedure is general because it includes sampling plans based on the minimax criterion and exact sampling plans in the special. cases.     

Demixing of Materials under Chemical Potential Gradients

A steady-state demixing of initially homogeneous solid solution which occurs in binary oxides under oxygen partial pressure gradient is treated by the path probability method of irreversible statistical mechanics from an atomistic point of view. The larger the difference between the diffusion coefficients of constituent species is, the larger is the degree of demixing. Essential features of demixing by diffusion, however, do not depend on whether the driving force is the chemical potential gradient, the electric field, or others.     

Demixing of Oxides under a Temperature Gradient

The microscopic mechanism of a demixing of binary oxide solid solution in an oxidizing atmosphere under a temperature gradient is investigated by the path probability method of irreversibly statistical mechanics. The problem is reduced to a steady-state metal ion transport under the coexistence of an oxygen partial pressure gradient and a temperature gradient. Calculations are performed under two different conditions: (i) demixing in a steady-state diffusion process under the two driving forces and (ii) a redistribution of constituent metal ions in a closed system under a given temperature gradient (Soret effect). The results can be interpreted as the superposition of the contributions of two driving forces plus a cross term which tends to zero as the driving forces become small. The difference between these two cases is discussed in detail and some controversies in the past work with respect to the difference are pointed out.