Taguchi Concepts

Figure 14.8 Taguchi concepts of inner and outer arrays. See text for discussion.

Chapter 10 presents three novel safety assessment and decision making approaches. They are (1) a safety based decision support systeni using artificial neural network techniques, (2) a safety optimisation framework using Taguchi concepts, and (3) A multiple criteria decision making approach applied to safety and cost synthesis. Such approaches provide the safety analyst with more flexibility to facilitate risk modelling and decision making. 

Keywords Analytical Hierarchy Processing, artificial neural networks, cost, decision making, decision support systems, marine safety assessment, multiple criteria decision analysis, risk assessment, Taguchi concept. 

Taguchi Concepts and Their Applications in Maritime Safety Assessment 10.2.1 Introduction... 

A Safety Optimisation Framework Using Taguchi Concepts... 

A safety optimisation framework using Taguchi concepts for maritime safety engineering applications presented in this Section has the following steps (Sii (2001), Sii et al. (2001)) ... 

Application of Taguchi Concepts in Maritime Safety Studies... 

Taguchi uses fractional factorial designs to determine the first-order effects of both the control factors and the noise factors, but he separates the factors and the designs into an inner array (involving the control factors only) and an outer array (involving the noise factors only). The concept is shown geometrically in Figure 14.8 for three control factors and three noise factors. 

Taguchi. He has developed both a philosophy and a methodology for the process of quality improvement, which depend heavily on statistical concepts and tools. Many Japanese firms have applied these methods with great success. 

Taguchi s ideas can be separated into two fundamental concepts, which are ... 

Because Taguchi has made no immediate coupling between the concept of loss-functions (the more philosophical part) and the off-line quality control methods (the more concrete part), the loss functions probably serve best as an important background concept rather than a practical working tool. 

In Taguchi methods, we define the quality of the product in terms of the deviation of some response parameter from its target or desirable value. The concept of quality and the idea behind the Taguchi philosophy is illustrated with the example shown in Figure 8.20. 

The basic concepts and seminal work in this field are from Taguchi, who stated that any product whose performance characteristics are different from the target values suffers a loss in quality, which he quantified by a parabolic function. He then classified factors as 1) control factors, which can be controlled under normal operating conditions and 2) noise factors, which are difficult, impossible, or very expensive to control. 

It is this requirement for robustness that prompted Taguchi to develop the concept of parameter design and produced a great impact on the world of experimental design. 

Concepts of Parameter Designs (Barker 1990 Taguchi and Wu 1980 Taguchi 1986)... 

Much of this work focuses on mean response and ignores the response variabihty. In order to improve processes, we would like to have processes with both desirable results and a minimum variation. This topic can be addressed using the quahty loss function concept suggested by Taguchi. 

Gaitonde, V.N. and Karnik, S.R. (2007) Taguchi robust design for multi-response drilling optimization to minimize burr size using utility concept. International Journal of Manufacturing Research, 2(2) 209-24. 

RD techniques based on how to exploit the mean and variance information of responses. The concept of building quahty into a design is increasingly popular in pharmaceutical industry because of their practicality. There have been many attempts to integrate Taguchi s RD principles with well-established statistical techniques in order to model the response directly as a function of control factors (Shin and Cho, 2005). In practice, it is necessary to handle the responses sampled as time series data that called time-oriented response in this paper. The associated experimental format is shown in Table 1. x, y, s, and t represent the vector of control factors, mean, variance, and time index respectively. 

---