Premium
Designing real time model mobile monitoring system for model predictive control in a nonlinear continuous stirred tank reactor
Author(s) -
Djarum Danny Hartanto,
Ahmad Zainal
Publication year - 2020
Publication title -
asia‐pacific journal of chemical engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.348
H-Index - 35
eISSN - 1932-2143
pISSN - 1932-2135
DOI - 10.1002/apj.2430
Subject(s) - continuous stirred tank reactor , pid controller , control theory (sociology) , model predictive control , response time , nonlinear system , process (computing) , robustness (evolution) , control engineering , controller (irrigation) , process control , engineering , computer science , temperature control , control (management) , physics , quantum mechanics , chemical engineering , artificial intelligence , agronomy , biochemistry , chemistry , computer graphics (images) , biology , gene , operating system
A continuous stirred tank reactor (CSTR) plays a major role in many process industries, and its control strategy could be very critical in achieving desirable yields, conversion, rates, and minimizing the formation of side products. CSTR control over a wide operating range has been a major issue to control engineers as it exhibits a highly nonlinear behaviour and time varying behaviour. A traditional control mechanism like a proportional–integral–derivative (PID) controller although still widely used, may not be sufficient in achieving optimum performance and robustness. In this study, model predictive control (MPC) strategy is developed for a nonlinear dynamic CSTR process, and the performance was compared with the traditional PID controller. The result shows that for a small set point change, MPC has similar performance to the PID controller. MPC performed much better when subjected to a much larger set point change in which the response time is 12 times faster and is able to reduce up to 60% of the loop interaction. The second part of this study is to design a real‐time mobile monitoring system (proof of concept). The overall design allows for real‐time updates on the performance of CSTR remotely. Engineers will be notified when the equipment fails to run at its full performance and generate critical analysis based on the historical data of the equipment. Lastly, a real‐time safety alert allows accidents to be predicted ahead before it happens and allows workers to be notified in real time.