Partial Least Square-based Model Predictive Control for Large-scale Manufacturing Process with Disturbances Partial Least Square-based Model Predictive Control for Large-scale Manufacturing Process with Disturbances

The Model Predictive Control (MPC) method has been widely accepted as a useful tool for keeping the quality on target in the manufacturing process. However, the conventional MPC methods are inadequate for the large-scale manufacturing process particularly in the presence of disturbances. The goal of the paper is to suggest a Partial Least Square (PLS)based MPC methodology for accommodating the characteristics of the large-scale manufacturing process with disturbances. The detailed objectives are: (1) to identify a reliable prediction model that handles the large-scale “short and fat” data, (2) to design an effective control model that both maximizes the required quality and minimizes the labor costs associated with changing the process parameters, and (3) to develop an efficient optimization algorithm that reduces the computational burden of the large-scale optimization. The case study and experimental results demonstrate that the presented MPC methodology provides the set of optimal process parameters for quality improvement. In particular, the quality deviations are reduced by 99.4%, the labor costs by 84.2%, and the computational time by 98.8%. As a result, the proposed MPC method will save cost and time in achieving the desired quality for the large-scale manufacturing process with disturbances.