pp. 2471-2480
S&M2631 Research Paper of Special Issue https://doi.org/10.18494/SAM.2021.3309 Published: July 15, 2021 Modified Gauss–Newton Algorithm for Evaluation of Full Lightning Impulse Voltage Parameters [PDF] Panuwat Jamroen, Poom Yoosorn, Suthasinee Waengsothorn, and Peerawut Yutthagowith (Received January 29, 2021; Accepted June 16, 2021) Keywords: curve fitting technique, Gauss–Newton method, Levenberg–Marquardt algorithm, lightning impulse voltage, two-exponential function
This paper presents an efficient curve fitting technique based on a Gauss–Newton (GN) method with a varied damping factor for the evaluation of lightning impulse voltage parameters. The proposed method can be used to determine the base curves of full lightning impulse voltages with oscillations corrected from the parameters of test cases according to the standard (IEC 61083-2). To verify the performance of the proposed method, base curves reconstructed from the developed method were compared with curves reconstructed from nonlinear least squares regression using the Levenberg–Marquardt (LM) algorithm, as prescribed in the standard. The proposed method provides the same results as the standard recommended method but its execution time is significantly shorter. The results of this study indicate that the proposed method is suitable for evaluating lightning impulse voltage parameters.
Corresponding author: Peerawut YutthagowithThis work is licensed under a Creative Commons Attribution 4.0 International License. Cite this article Panuwat Jamroen, Poom Yoosorn, Suthasinee Waengsothorn, and Peerawut Yutthagowith, Modified Gauss–Newton Algorithm for Evaluation of Full Lightning Impulse Voltage Parameters, Sens. Mater., Vol. 33, No. 7, 2021, p. 2471-2480. |