FOPDT modelling and controller comparative study for smart tube aqua filter (STAF) / Ahmad Aftas Azman ...[et al.]

Azman, Ahmad Aftas and Mohammad, Nurul Nadia and Sidek, Norbaya and Ali, Mohd Fozi and Abu Bakar, Ilyani Akmal and Taib, Mohd Nasir and Rahiman, Mohd Hezri Fazalul (2018) FOPDT modelling and controller comparative study for smart tube aqua filter (STAF) / Ahmad Aftas Azman ...[et al.]. Journal of Electrical and Electronic Systems Research (JEESR), 12 (1): 10. pp. 66-73. ISSN 1985-5389
Abstract

Malaysian Meteorological Department statistics stated that rainfall in this country is among the highest the world with 2500mm a year. Therefore, the phenomenon of flooding often occurs in Malaysia. Flooding will cause water to become cloudy or turbid. This situation is getting worse as demand for fresh water is higher than supply. Therefore, researchers have designed several of water filter system to overcome these issues. Not to mention, the usage of the chemical as a purifying agent such as coagulant and flocculation process for turbidity water purification. In this study, we used Aluminium Sulphate Al 2 (SO4) 3 as a coagulant agent, and Smart Tube Aqua Filter (STAF) plan in flocculation process. This plant is modeled by first order plus time delay (FOPDT) model, to explain the behavior of processes. Model Predictive Controller (MPC) and proportional integral derivative (PID) controller are applied to control the process. Robustness and efficiencies of the controller are analyzed and evaluated, and then a comparative study is conducted for the best performance of transient response. Finally, we found that MPC gives the best performance of transient response in robustness test (step test has faster settling time, of 5035.0 sec and overshoot percentage of 0.0917%, Setpoint change test has performed better in five (5) categories of the test), compared to PID controllers for STAF simulation test.

Item Details
Edit Item
Edit Item
Downloads & Files
[thumbnail of 63047.pdf]
Text
63047.pdf
Download (518kB)
Indexing & Metrics