Akademik Veri Yönetim Sistemi
Otomatik Kontrol Ulusal Toplantısı, TOK 2023, İstanbul, Türkiye, 14 - 16 Eylül 2023, ss.1-6
In this study, the control performances of proportional-integral controllers developed for single-phase uninterruptible power supply inverters whose parameters are optimized using the artificial bee colony algorithm are analyzed under different load conditions. The nonlinear dynamic system model of the inverter makes it difficult to tune the controller parameters with classical design methods. It is inevitable that controllers developed using a linear system model which exhibits acceptable performance for a given operating point and the near it are inadequate in performance indices such as reference tracking and total harmonic distortion. Heuristic algorithms such as the artificial bee colony algorithm analyze the performances of possiblecontroller parameters at different operating points of the inverter and determine the controller parameters that provide the best performance. In this work, total absolute error and total squared error are used as cost functions, and the performances of two different PI controllers with tuned parameters according to those two functions are analyzed. The PI controller, whose parameters were adjusted utilizing the classical method, was used for comparisons. The results of the simulation studies carried out in Matlab environment show that the controller developed using the cost function of integrated absolute error provides an improvement in the total harmonic distortion performance index.
In this study, the control performances of proportional-integral controllers developed for single-phase uninterruptible power supply inverters whose parameters are optimized using the artificial bee colony algorithm are analyzed under different load conditions. The nonlinear dynamic system model of the inverter makes it difficult to tune the controller parameters with classical design methods. It is inevitable that controllers developed using a linear system model which exhibits acceptable performance for a given operating point and the near it are inadequate in performance indices such as reference tracking and total harmonic distortion. Heuristic algorithms such as the artificial bee colony algorithm analyze the performances of possible
controller parameters at different operating points of the inverter and determine the controller parameters that provide the best performance. In this work, total absolute error and total squared error are used as cost functions, and the performances of two different PI controllers with tuned parameters according to those two functions are analyzed. The PI controller, whose parameters were adjusted utilizing the classical method, was used for comparisons. The results of the simulation studies carried out in Matlab environment show that the controller developed using the cost function of integrated absolute error provides an improvement in the total harmonic distortion performance index.