Open AccessTemperature and water level control in a multi-input, multi-output process using neuro-fuzzy controller
Author(s) -
M V Akbariza,
Djati Handoko,
Prawito Prajitno
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1816/1/012022
Subject(s) - control theory (sociology) , settling time , controller (irrigation) , overshoot (microwave communication) , pid controller , temperature control , volumetric flow rate , decoupling (probability) , mean squared error , mathematics , process control , process (computing) , environmental science , computer science , control engineering , engineering , step response , statistics , mechanics , physics , control (management) , artificial intelligence , telecommunications , agronomy , biology , operating system
In this research, a simulation study for temperature and level control in a liquid (water) mixing process is proposed using MATLAB/Simulink. The objective of this control system is to maintain the temperature and water level at the set points in a liquid mixing process by controlling the flowrate of cold and hot water that enters the mixing tank. The mixing tank used in this study is assumed to have a volume of 80 liter, while the maximum flowrates of both water inputs are 15 liter/min, and the maximum temperature and level in the mixing tank are 90 C and 75 cm, respectively. The influence of one variable to the other is reduced using decoupling technique. In the development process of the controller, PI controller is used to generate the training data required by the ANFIS-based controller. The performance of the proposed controllers has been tested with several set points changes by observing its performances parameters, such as RMSE, rise time, settling time, and % overshoot as quantitative data. It also has been compared with a PI controller using the same set point changes as the ANFIS-based controller did. These results show that the ANFIS-based controller is generally better than the PI controller. It has the average RMSE values of 0.174 and 0.196 for temperature and level control respectively, while the PI controller has 0.21 and 0.20 average RMSE values for temperature and level control.