A transient period control methodology for continuous mix manufacturing.

Abstract

The startup of a manufacturing system presents a transient period where the process is not in statistical control. Many traditional process control methods rest on an underlying assumption that the system is already in steady state and thus they are confounded by the disturbance mechanism associated with initiating a complex system. The focus of this dissertation is on transient period adjustment of the continuous mix manufacturing system because manufacturing companies incur large losses during the transient period due to out-of-specification product and because the current literature does not directly address the issue. In this dissertation, we first formulate the continuous mix manufacturing problem in a way that permits analysis of the transient period. Then, we develop a multiple objective nonlinear control function (MONCF) to regulate the system. The control function addresses both management concerns and operating conditions. Management concerns include economic cost containment and finished product quality specifications. Operating conditions include relationships between process variables that can be predicted via statistical analyses before manufacturing actually commences. The control function is a mathematical model which we solve using goal programming and nonlinear optimization. To advance the control function presented here, we develop a transient period process adjustment methodology (TPPAM). Due to the changing nature of the transient period, there is a necessity provide dynamic approach rather than rest with one static optimal solution. The TPPAM guides the control function for startup conditions and real-time operation of the system. In this way, we move beyond some initial "recipe" of setpoints to adjusting setpoint decisions based on the knowledge that a transient period will precede the steady-state period. A part of the challenge in this work is to provide a methodology that is solidly within reach of manufacturing practitioners. Therefore, we emphasize the use of controllable decision variables and measurable response variables. An example from a case study operation is used to demonstrate and validate the methodology. Results using the case study indicate that the TPPAM provides a reduction in scrap. Further, the length of the transient period is reduced, allowing for greater flexibility in production scheduling.