International Journal of Applied Science and Technology

ISSN 2221-0997 (Print), 2221-1004 (Online) 10.30845/ijast

Estimation of the Confidence Intervals for the Average Loss Function of Taguchi and Signal to Noise ratio through Resampling Bootstrap Method
Marcos Augusto Mendes Marques, Anselmo Chaves Neto

The constant improvement in the manufacture of products is a way for these companies stay competitive in the market, and for this to be possible, investments in quality become essential. Such improvements aim for the production of more uniform products and robust in front of functional variations. To measure the uniformity and robustness of those products is necessary the use of appropriate tools, being these tools (quality indicators) the average loss function of Taguchi and the signal to noise ratio. Through the average loss function of Taguchi it can be measured the average loss, in monetary units, due to functional variation of a variable answer regarding its project (target), being this quality indicator used to justify investments in quality. The signal to noise ratio assists, in the project phase, in the fixation of values of parameters with the intention of minimizing the effect of noises (undesirable and uncontrollable factors) in the performance of some characteristic of the product, synthesizing in a single value the relationship between the wanted sign and the unwanted (noise). This way, such indicators justify the importance of investments in quality searching for, like this, centered productive processes and uniforms. Due to its contribution in the area of quality engineering, it becomes interesting the accomplishment of statistical inferences in the two quality indicators mentioned, just as the estimate of confidence intervals. As a classic statistical method does not exist to estimate of these intervals, is proposed to do that, their estimates are done through the resembling method computationally intensive, due to the generality of its application and reliability of the results.

Full Text: PDF