A KL/Subspace Modeling Method for Integrated Design and Control in Curing Processes

In this Article, a data-driven modeling approach is proposed for hybrid design/control variables in complex curing ovens. The coupling influence of the design/control variables is first separated into two subtasks: modeling respectively for the design-variable-dependent basis function and the control-variable-dependent temporal model. Each subtask only considers the nature of its corresponding variable and does not interact with the other variable, which will be less complex and more easily modeled. The original system is then reconstructed by synthesizing these two submodels. The advantage of this proposed method is that the well-developed design methods and control methods can respectively handle the design-variable-related model and the control-variable-related model, which will benefit to achieve the desirable overall curing performance by integration of design and control. Finally, the effectiveness of the proposed method is verified by the modeling of an actual curing oven.