Mohammad Mirabi; Mohammad Taghi Fatemi Ghomi; Fariborz Jolai
Abstract
The design of control chart has economic consequences that pure statistical viewpoint does not consider them. The economic-statistical design of control chart, attends not only statistical properties such as average time to signal (ATS) but also economic consequences like hourly expected total cost. ...
Read More
The design of control chart has economic consequences that pure statistical viewpoint does not consider them. The economic-statistical design of control chart, attends not only statistical properties such as average time to signal (ATS) but also economic consequences like hourly expected total cost. The x-bar control chart dominates others if the quality is measured by continuous scale. This paper has considered the economic-statistical design of variable sample size and sampling interval (VSSI) x-bar control chart with multiple assignable causes. Using three sample sizes and three sampling intervals to construct the VSSI x-bar control chart and considering possible combination of design parameters as a decision-making unit, are part of novelty of this research. The problem is formulated as multiple objective decision making (MODM). Also, one capable hybrid meta-heuristic based on genetic algorithm is developed in this research and it was compared with some approaches extracted from the literature and it is found that it can be competitive based on economic and statistics factors.
Ali Salmasnia; Ali Tavakoli; Maryam Noroozi; Behnam Abdzadeh
Abstract
Control charts are powerful tools to monitor quality characteristics of services or production processes. However, in some processes, the performance of process or product cannot be controlled by monitoring a characteristic; instead, they require to be controlled by a function that usually refers as ...
Read More
Control charts are powerful tools to monitor quality characteristics of services or production processes. However, in some processes, the performance of process or product cannot be controlled by monitoring a characteristic; instead, they require to be controlled by a function that usually refers as a profile. This study suggests employing exponentially weighted moving average (EWMA) and range (R) control charts for profile monitoring, simultaneously. For this purpose, the parameters of these control charts should be determined in a way that the expected total cost is minimized. In order to evaluate the statistical performance of the proposed model, the in-control and out-of-control average run lengths are applied. Moreover, the existence of uncertain parameters in many processes is a barrier to attain the best design of control charts in practice. In this paper, the economic-statistical design of control charts for linear profile monitoring under uncertain conditions is investigated. A genetic algorithm is used for solving the proposed robust model, and the Taguchi experimental design is employed for tuning its parameters. Furthermore, the effectiveness of the developed model is illustrated through a numerical example.