Numerical Inverse Fourier Transform for Accurate Electromagnetic Transient Analyses

Abstract

This work presents a novel numerical inverse Fourier transform for electromagnetic transient analysis emphasizing the lightning phenomena. The inverse Fourier integral is separated into nonuniform frequency intervals, and exponential interpolation is employed for each interval before applying analytical integration to acquire a continuous transient function in the time domain. In addition, a nonuniform frequency sampling approach was proposed to reduce the number of frequency samples. The performance of the proposed technique was verified using some example cases and compared with the previously developed methods, i.e., conventional inversed fast Fourier and linearizing Fourier integral techniques. In terms of computational efficiency and accuracy, the proposed technique is an alternative and effective tool for frequency-domain-based transient analyses.