IMPROVING THE PERFORMANCE OF BASIC PROCESS CONTROL SYSTEM FOR OPTIMAL CEILING BOARD PRODUCTION

This paper “improving the performance of basic process control system for optimal ceiling board production” was presented for the intelligent control of dynamic process. The study utilized Feed Forward Neural Network (FFNN), trained with the data of the ceiling board plant collected at +3% (200 liters) of fluid stability point and used for the adaptation of Programmable Logic Controller (PLC). The FFNN was trained with back-propagation algorithm and evaluated using Mean Square Error (MSE) and Regression (R). The MSE performance after training was 3.8417e-10, while the R was 0.97173. The FFNN was integrated on the PLC and used for the control of the ceiling board plant, using simulation methodology. The result showed that the chemical fluid stability was maintained at tolerance of +3% (200 liters).
Keywords: Programmable Logic Controller; Feed-Forward Neural Network; Ceiling Board