Feedback Control and Statistical Process Monitoring

In the industrial world, different methodologies have been applied to reduce variability. Engineers whose training is in the quality assurance sciences have frequently used statistical process control (SPC) whereas engineers whose background includes control theory have employed engineering process control (EPC), often in the form of feedback control. SPC is usually applied to processes that vary about a fixed mean, with successive observations viewed as statistically independent. EPC is usually applied to processes in which successive observations are related over time, and where the mean drifts dynamically. The objective of SPC is to reduce variability by the elimination of assignable causes that shift the mean off-target, while EPC continuously adjusts manipulatable variables in the process to keep the mean on target. This study demonstrates the effectiveness of integrating SPC with feedback control for three commonly encountered schemes: Integral Control, Proportional-Integral Control, and Proportional-Integral-Differential Control. A simulation study demonstrates that the combined EPC/SPC scheme significantly reduces overall variability in comparison to feedback control alone when assignable causes are present.