Spatial decomposition as a tool for the evaluation and design of cross-directional control systems in sheet metal rolling

Over supply in many segments of the metal strip manufacturing industry drives a need to achieve improved product quality and yield from existing mill stock. This demands improved process control, which is achieved though a range of commercially available actuators, most of which can be used on existing mills. However, these actuators are expensive and frequently their effect on product quality cannot be predicted prior to installation. This paper uses the results of cross-directional control theory to propose a rationale for evaluating such decisions. It demonstrates that existing quality error signals can be used to define actuator requirements and gives a means of evaluating the performance of existing control schemes.