An Automatic Frequency Controller for Kotmale Hydro Power Station using Fuzzy Logic

During supply demand mismatches, the system frequency can either fall or rise compared with its previous steady state operating value. In such instances, the generator governors either increase or decrease the generation (in accordance with the governor droop setting) to restore the supply demand balance and stop the ensuing frequency change (fall or rise). However, even after the supply demand balance has been restored and the frequency has reached a new steady state value, there can be a frequency error. The governor action assists to minimize the steady state frequency error. However, zeroing the frequency error and bringing the system frequency back to the rated value need adjustment to the "speed reference" of governors. Although this is normally done by Automatic Generation Controllers (AGC) that are on secondary load-frequency control, in Sri Lanka, the frequency controlling act is performed entirely manually through operator intervention. Such manual frequency control has a number of draw backs, including possible operator errors and longer response times. One of the four power stations designated to perform the frequency control function in the Sri Lankan power system is at Kotmale. This paper presents an automatic frequency controller using Fuzzy Logic to carry out frequency controlling at Kotmale power station. The Fuzzy Logic based controller is designed to replace the action of the operator, who presently controls the frequency manually. The paper presents the modeling and the simulation of the existing frequency control system using relevant parameters of the Kotmale governor and the chronological load curve of the Sri Lankan power system. MATLAB/Simulink (TM) 2013 software package is used to calculate the steady state frequency error. Subsequently, the parameters are fine-tuned and the Fuzzy Logic Controller (FLC) is introduced as a secondary loop frequency controller and the study is repeated. Mamdani type fuzzy inference system is employed here to implement this project because of its simplicity, robustness, reliability and intuitive characteristics. Finally the simulation results using the FLC are compared with the existing frequency controlling system of Kotmale power station to prove that the FLC yields improved control performance.