COMPARISON PID CONTROLLER TUNING METHODS ON BOILER DRUM LEVEL CONTROL SYSTEM OF PLTU

• Main Author: AZHAR HARIYADI (NIM : 23814306), DODY
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/26744
• Institution: Institut Teknologi Bandung

<p align="justify">Drum-Boilers are the source of the 'hot' steam that a steam turbine requires to produce in a Steam Power Plant (PLTU). There needs to be a water level equilibrium in the boilers to make the steam production process inside perfect. If the level is too low, the tubes in the boiler will break due to overheating. Whereas if the water level is too high causes too much water content in hot steam that can damage the blade of the turbine. <br /> <br /> <br /> In this research, case is taken on the drum-boiler unit-1 PLTU Pelabuhan Ratu. The study aims to model the system, then determine the PID controller parameter values in accordance with the characteristics of the system obtained from the modeling results. Modeling is done in some areas of the drum-boiler system that refers to the workload of the plant. This is done because of system nonlinearity and requirements on existing PID controllers. The modeling is done by using Box-Jenkins model structure with PEM (Prediction Error Methode) estimator. The tuning of PID controllers is done by the Continuous Cycling methods of Ziegler-Nichols and Tyreus-Luyben, and compared with the optimization method using Genetic Algorithm. Genetic Algorism Methods are used to look for minimum error using ISE (Integral Square Error), IAE (Integral Absolute Error) and ITAE (Integral Time Absolute Error) fitness functions on the system. The tuning method with IMC (Internal Model Control) is also used as a comparison. <br /> <br /> <br /> The result of modeling to 4 area shows RSME value between 5.1 ~ 7.0 which shows the model represents the dynamics of drum-boiler well. The result of tuning PID controller with 6 methods above is then simulated with drum-boiler system and given disturbance in the form of input noise and output noise. The result of the tuning of the controller by the error optimization method can make a good controller for the system in all areas, while the Continuous Cycling method Zigler-Nichols Tyreus Luyben and IMC is only able to make the system stable in some areas only. Among the three methods of optimization of the error, the use of ISE objective function criteria shows the best performance compared to the same method with the objective function of IAE and ITAE<p align="justify"> <br />