Real-Time Optimization using Gradient Adaptive Selection and Classification from Infrared Sensors Measurement fo Esterification Oleic Acid with Glycerol

Abstract

Purpose- The production of glycerol derivatives by esterification process is subject to many constraints related to the yield of production and efficiency. The purpose of this paper are to propose a real-time optimization using gradient adaptive selection and classification form infrared sensor measurement to cover various disturbance and uncertainty in the reactor. Design/methodology/approach- The integration of esterification process using Self-Optimization (SO) was developed with classification process was combined with Necessary Condition Optimum (NCO) as gradient adaptive selection supported with laboratory scaled medium Wavelength Infrared (mid-IR) sensors, and measured the proposed optimization system indicator in batch process. A Business Process Modeling and Notation (BPMN 2.0) was built to describe the task of SO workflow in collaboration with NCO as an abstraction for conceptual phase. Next, Stateflow modeling was deployed to simulate the three state of gradient-base adaptive control combined with Support vector Machine (SVM) classification and Arduino microcontroller for implementation. Findings- This new method shows that the responsiveness of control increase product yield up to 13%, lower error measurement with percentage error 1.11%, reduce the process duration up to 22 minutes, with an effective rang of stirrer rotation set between 300 to 400 rpm and final temperature between 200 to 210 C which was more efficient, as it consumed less energy. Practical Implementation- This research introduces a new development of RTO implementation to optimal control and as such marks the starting point for more research of its properties. As the methodology is generic, it can be applied to different optimization problems for batch system in chemical industries. Originality- The paper presented is as original as it present the first application of adaptive selection based on gradient value of mid-IR sensor data, applied to real-time determining control state by classification with SVM algorithm for esterification process to increase the efficiency.