Lightning Impulse Parameter Estimation Using Nonlinear Least Squares Algorithm

This paper presents a new technique based on nonlinear least square (NLSQ) algorithm for estimating the mean curve and parameters of the standard lightning impulse (LI) waveform as required by IEC 60 and IEEE Std. 4. The proposed NLSQ technique is applied to analytical LI waveforms to estimate the mean curve of the impulse using a bi-exponential model after removing high frequency components from the simulated (measured) impulse. The effectiveness of the proposed technique is assessed by conducting several cases of LI impulse waveforms with front and tail oscillations and LI waveform with overshoot. The results of the NLSQ technique are compared with those obtained by the particle swarm optimization (PSO) algorithm. The simulation results show that the proposed NLSQ technique is robust and effective in estimating the peak value, front and tail times of the impulse waveform accurately.