A TLBO optimized photovoltaic fed DSTATCOM for power quality improvement

A common practice of obtaining the values for filter parameters and gains of proportional plus integral (PI) controller is to use a linear model of the distributed static compensator (DSTATCOM). The values so obtained produces perfect harmonic cancellation (PHC) and unity power factor (UPF) for balanced and sinusoidal supply voltages based photovoltaic fed DSTATCOM (PV-DSTATCOM). However when supply voltages are distorted; PV-DSTATCOM may not produce PHC along with UPF due to non-linear behavior of the passive filter parameters. Hence; there should be an optimal calculation for filter parameters and PI controller gains; so that the PV-DSTATCOM can deliver optimal performance under distorted supply voltages. This paper presents a new algorithm based on teaching-learning behavior inside a classroom; to optimize the gains of PI controller and filter parameters. Through simulation results; it is observed that the optimization based control algorithm performs quite satisfactorily under sinusoidal as well as distorted supply voltage conditions. The proposed teaching-learning optimization based PV-DSTATCOM is compared with grenade explosion method (GEM) and found to converge faster than GEM to reach the global optimum solution.