Hamid Rahimi Esfahani - Department of Electrical EngineeringIslamic Azad University, Najaf Abad Branch,
Rasoul Amirfattahi - Digital Signal processing Research Lab., Department of Electrical and Computer Engineering
Farshad Kiyoumarsi - Department of Computer Engineering,Islamic Azad University, Shahrekord BranchShahrekord, Iran

In this paper, among different methods of signal processing, wavelet transform is chosen due to its advantages over other methods. To show wavelet transform capabilities, first an HVDC system which has noise on its output current is simulated. In the first step noise is removed by applying discrete wavelet transform. In the next steps, harmonic problem is resolved thorough appropriate passive filters. In this paper, we suppose that not only low order harmonics exist in the output current, but also high order ones. The results indicate that we can obtain a good passive filter design for harmonic reduction by decomposing a signal into its harmonic components via applying discrete wavelet transform. For more confirmation of this method, the scalogram of wavelet coefficients is presented. At the end, to compare this method with traditional one, the figures of traditional method for passive filter designing is presented

multi-resolution decomposition; wavelet transform; passive filter for harmonic reduction; harmonic detection; wavelet-based threshold de-noising method