In recent years, direct financing such as bond issuing has become common to many Japanese firms. The pricing method for the corporate bond differs from that for risk free bonds such as JGB, since corporate bonds incorporate credit risk. On default, none or only a part of principal and interest is paid to bond holders. So, it is necessary to evaluate the credit risk when investors price bonds which embed credit risk. The use of credit rating to evaluate credit risk is quite popular with many researchers. The pioneering paper by Jarrow et al.(1997) models the probability of credit rating change as a Markov chain. They make possible the pricing of corporate bonds using the following procedure. First they define martingale default probability, and then calibrate the martingale default probability to bond price. In their paper, they give the recovery rate at the default endogeneously. But in Japanese bond markets, issue of corporate bond was rare until quite recently. So, few defaults have been observed and statistical parameter estimation is quite difficult. In this paper, we propose a procedure to estimate the recovery rate. Then we conduct an empirical study in Japanese bond market and show some findings.

View full abstract

The work quality and efficiency of self-threading screw tightening operations, in a three-dimensional working space at a standing position, is affected by the working postures. Empirical experiments are designed to investigate those effects. The following findings are identified from results of experiments : In some work directions and locations, tightening torque was not stable and went outside of the allowable range; Adequate tightening torque could not be obtained in such operations as 1) left-directed operations at the right of the body, 2) forward-directed operations at the far side of the body, and 3) downward-directed operations at high locations; The maximum vertical pressure and the maximum horizontal power waste for a driver were not stable in the work directions and locations; Excess pressure could be seen in such operations as 1) left-directed operations at the front of the body and 2) forward-directed operations at most locations; Bigger horizontal power waste could be seen in such operations as 1) left-directed operations at most locations and 2) forward-directed operations at low locations. These effects on work quality and efficiency must be considered in the work design and the product design.

View full abstract

An EIQ table analyzes the relationship between items and shipments in distribution centers and warehouses. This method is used for the planning of distribution centers and warehouses for the selection of the machines and equipment. The EIQ table is a useful technique for the analysis of logistics systems. The EIQ table is analyzed using the method of descriptive statistics. The analysis employs mean value and standard deviation, so it is difficult to carry out a quantitative analysis of warehouses and distribution centers. It is necessary to theoretically analyze characteristics of customers and goods with the EIQ table. This paper proposes the GM(1, 1) model for the analysis of the EIQ table. The procedure of the preparation of the EIQ graph is also explained. Gray theory is applied for accumulation of IQ data and accumulation of EQ data. Gray theory makes it possible to show the shape of the EIQ curve. The shape of the accumulation graph is analyzed from the change of the parameters. It is possible to classify the shape of the EIQ curve from the parameters of the GM(1, 1) model. The theoretical formula in search of the name of goods which came in at the ratio with the largest shipment in the IQ model is introduced. The relationship between a and u of the parameter is introduced from the theoretical formula. In addition, the equation for deciding the scale of the distribution center or warehouse is clarified. The design of a distribution center or warehouse is possible from the theoretical formula of GM(1, 1). The applicability of this paper as well as its effectiveness is illustrated by case studies.

View full abstract

Today, CAD is widely used in the design department. If the CAD information can be introduced into the image processing, the ability of recognition can be increased extremely. In this paper, we propose the recognition method by image processing using 2D or 3D CAD information. When 2D CAD information is used, an object is recognized as follows. An image of the object taken from one camera is represented by the characteristics, which are its center of gravity, contour, the distance between the center of gravity and contour and FFT. Then the object is recognized by comparing these characteristics with those of 2D CAD figures. When 3D CAD information is used, there are two methods. In the first method, 3D CAD figure is transformed into 2D CAD figures that are seen from some given viewpoints. Then an object is recognized by comparing 2D CAD figures with an input image of the object using the method of 2D CAD information. In the second method, coordinates of vertexes on an object are recognized by images taken from several cameras. 3D characteristics of the object, which are the distance among vertexes and the normal vector of a plane made by vertexes, are estimated by these coordinates of vertexes. Then the object is recognized by comparing these 3D characteristics with those of 3D CAD figures. A classification of these methods makes it easy to determine dimensions of CAD and an image processing when objects are recognized. Further, the effectiveness of the recognition method comparing 3D characteristics of the object with those of 3D CAD figures, is confirmed by experiments using an existing handling robot system.

View full abstract

We formulate an optimal service capacity allocation problem in tandem production systems, and propose a solution method via second order cone programming. Given total amount of service capacities (service rates), the problem is to find an optimal allocation of the capacities to the production stages so as to maximize the throughput in the steady state. Since it is difficult to know the exact value of the throughput, our approach is based on the so-called "sample-path optimization." It finds an approximate optimal allocation by maximizing an approximate value of the throughput obtained through a simulation run under a fixed series of random numbers. We show that the maximization problem can be converted into a second order cone programming problem with many linear constraints and a low number of second order cone constraints, and an effective algorithm developed. Numerical experiments show that our method can be applied to large size systems with more than 100 stages.

View full abstract

In automated machining systems, interruption in machining due to the sudden fracture of a cutting tool is detrimental for constructing fully automated machining systems. Not only does it affect the quality of the machined surface, but it also halts production. It is well known that tool-life is probabilistic in nature and shows considerable variations. A statistical approach based on tool-life distribution is well recognized for designing effect and economical automated machining systems. Production halt by sudden fracture of cutting tool results in the loss of available time and materials, and the increase of production cost. This paper presents statistical models to analyze the production costs for fully automated unmanned machining systems as well as for semi-automated manned systems. A new criterion based on production cost is proposed for shifting to the fully automated machining system from the semi-automated manned system. The shortage of machined products caused by the interruption in the fully automated unmanned system is supposed to be supplied by the stocked machined products. Supply costs of the shortage are formulated in production cost. A statistical decision procedure for shifting to the fully automated unmanned machining system from the semi-automated manned system is given.

View full abstract

Taking account of the market situation, we have taken up cyclamen, begonia, zygocactus and clivia for our consideration and examined the planning of optimum use of greenhouse bench space. The previous paper treated a case where the same production plan is assumed to be repeated every year. In recent years, due to the lowering of interest rates, many potted flower producers have built new greenhouses on the site of an old one or on a vacant lot in the same compound. In view of this trend, this paper has dealt with optimization of long-term planning of potted flower production for a period of several years after the construction of a new greenhouse. In this examination, we have paid special attention to efficient use of greenhouse bench space, profitability and growth properties of the flowers. Moreover, making use of the model we have constructed, we have also examined the way in which we should implement a shift from the current production system to an optimum one.

View full abstract

The Fuzzy-set Concurrent Rating method (FCR method) was developed to measure subjects' attitudes more naturally than traditional rating methods (Oda, 1995). The present paper proposes a new improved definition of the method so that it can include both recent and future research. With the new definition, the improved FCR method, which is composed of a convenient set of procedures is proposed. For the purpose of assisting the users to understand it, the algorithm of integration using Mamdani's "min-max-gravity method" is explained in detail. The integration formulae are presented as well. The formulae can fit data from 2- to 5-item arrangements by employing Mamdani's "min-max-gravity" method and Mizumoto's "algebraic product-sum-gravity" method. Besides, a variety of procedures for the application of the FCR method to psychological measurements are proposed. After the proposal of the procedures, the guidelines to apply them to practical problems are shown in table form.

View full abstract