ELEKTRIKA - UTM Journal of Electrical Engineering, Vol 13, No 2 (2011)

Font Size:  Small  Medium  Large

Artificial Neural Network Controlled Multilevel Inverter based Active Power Filter for High Voltage Systems

Madhukar Maruti Waware, Pramod Agarwal

Abstract


In this paper, a five-level multilevel inverter based shunt active power filter is presented. In the industry, use of High Voltage (HV) nonlinear load is increasing considerably and elimination of harmonics in HV system is gaining  importance. Implementation of multilevel inverter for active power filter (APF) in HV system eliminates use of bulky and high cost transformer. The reference signal is generated by using the average power method instead of p-q theory to successfully eliminate harmonics of the distorted supply voltage. Carrier phase shifted pulse width modulation (CPS-PWM) scheme is implemented to generate switching signals which reduces the individual device switching frequency. The artificial neural network (ANN) is trained with conventional controller data to stabilize DC side voltage accurately and quickly as compared with  conventional controller over varied loading conditions. The simulation results have been validated with experimental results to  prove the effective elimination of harmonics using multilevel inverter in HV system.


Full Text: PDF

Refbacks

  • There are currently no refbacks.


ISSN 0128-442

ELEKTRIKA is now included in the:
Directory of Open Access Journals (DOAJ)