A Study of Fuzzy Based Method Synchronous Generator Terminal Voltage Control

Abstract

The voltage stability and power quality of the electrical system depend on proper operation of Automatic Voltage Regulator (AVR). Nowadays, the design technology of AVR is being broadly improved. Nonlinearities, parametric uncertainty, ill- defined mathematical model are an evitable problem faced in controlling the output voltage of Synchronous Generator (SG) leading to greater complexities in the design of the control system. Therefore, the application of Artificial Intelligence based controllers in electric power systems is becoming an important field of research. In the present work, conventional exciter type (IEEE DC1) and intelligent controllers have been suggested to replace the excitation circuit for improving the dynamic performance of the AVR SG system. The performance of the two controllers has been compared and assessed at different machine loading conditions. In this thesis, the performance comparison between two suggested controllers is based on how well these controllers improve the dynamic responses of SG when exerting to different loading conditions and different durations of fault application. The results show that the intelligent controller can give better dynamic behavior than its competitor. Moreover, when applying worst fault condition (short circuit case), it is seen that the fuzzy logic (FL) controller can keep satisfactory and stable dynamic characteristic to longer fault exertion time interval. The effect of changing the FL controller parameters on the dynamic performance of SG has been considered. The type of MF, number of MFs and knowledgebase elements are selected as the design parameters for this assessment. For specified knowledge-based elements and fixed number of MFs, the performance of intelligent controller has been evaluated based on type of MF and the duration length of load exertion. The simulation revealed that the FL controller based on triangular function could give best dynamic performance than that with other MFs. However, the FL controller based on Gaussian MF could give stable behavior to a longer fault period. The effectiveness of FL controller is also examined by different fuzzification methods. Mamdani and Sugenio fuzzification methods have been selected to synthesize the FL controller and performance comparisons are assessed in terms of ISE measure and the ability of controller to withstand longer fault exertion interval. The simulation shows that the performance of FL controller based on Sugenio method outperforms that based on Mamdani. The work also elaborates in detail the modeling and analysis of the Synchronous Generator that is connected to the steam-turbine. The modeling of Synchronous Generator and Steam turbine has been done in MATLAB/SIMULINK by integrating all the sub-models which include exciter, turbine, governor and the electrical part of Synchronous Generator.