DESIGN OF CONTROLLER AND OBSERVER FOR ONE TANK AND TWO INTERACTING TANK SYSTEM

• Main Author: Permata Sari, Hanisa
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/50284
• Institution: Institut Teknologi Bandung

System control process is important in the processing and production process. Tank level control system is one of the systems that is encountered in industrial process. There is demo set of dry tank emulator, which is the form of an electric circuit that represents the dynamics of the real tank system. Demo set can be used as a learning media for system monitoring with modelling and simulation. In a multi tank height control system, the component that potentially generate error is the field instrument and this error is harmful for the operation in a plant. In this final project, an observer will be designed on a one tank and two interacting tank systems to estimate the state of the dry tank emulator plant by providing a deviation in the instrument field, so the error value from these conditions can be known. Observer and controller have been designed in the system. Sensitivity test by applying ± 10% deviation on capacitor and resitot value have been done too see the system response. After that, the simulation is carried out by implementing an observer without controller and with PI controller for a single tank system and PID controller for a two interacted tank system. Of all those simulations, the worst conditions were selected for further analysis. The simulation results show that the estimated state of the observer state has been designed in a single tank system that represents the state of the plant with a MAE value of less than 1%. Meanwhile, the estimated state of the observer in the two- tank system is still not optimal and can be seen in the "worst case 1" MAE value between the PV capacitor C1 and PV capacitor C1 is 7.913%. The tuning method that has been used does not get the system response that is desired characteristics because it takes only calculate the location of the closed system pole and does not calculate the zero in the controller.