Problems of Special Causes in Feedback Adjustment

  • Lee Jae June (Department of Statistics, Inha University) ;
  • Cho Sinsup (Department of Statistics, Seoul National University) ;
  • Lee Jong Seon (Department of Statistics, Inha University) ;
  • Ahn Mihye (Department of Statistics, Inha University)
  • Published : 2004.04.01

Abstract

Process adjustment Is a complimentary tool to process monitoring in process control. Process adjustment directs on maintaining a process output close to a target value by manipulating another controllable variable, by which significant process improvement can be achieved. Therefore, this approach can be applied to the 'Improve' stage of Six Sigma strategy. Though the optimal control rule minimizes process variability in general, it may not properly function when special causes occur in underlying process, resulting in off-target bias and increased variability in the adjusted output process, possibly for long periods. In this paper, we consider a responsive feedback control system and the minimum mean square error control rule. The bias in the adjusted output process is investigated in a general framework, especially focussing on stationary underlying process and the special cause of level shift type. Illustrative examples are employed to illustrate the issues discussed.

Keywords