Improved criteria determination of an automated negative lightning return strokes characterisation using Brute-Force search algorithm

• 第一著者: Abdul Haris, Faranadia
• フォーマット: 学位論文
• 言語: 英語
• 出版事項: 2021
• 主題: Lightning; Atmospheric electricity
• オンライン･アクセス: http://psasir.upm.edu.my/id/eprint/104072/1/FARANADIA%20BINTI%20ABDUL%20HARIS%20-%20IR.pdf

Over the years, many studies have been conducted to measure, analyse and characterise (i.e. mainly on the negative return strokes) the lightning-generated electric field waveform for a better understanding and conception of the lightning phenomenon and processes. The characterisation also contributes to the lightning protection system and lightning location system development as these systems require measured and validated data of the negative lightning return stroke characteristics. Previous studies mostly performed the characterisation of the negative return strokes using a conventional method, namely manual observations. Nevertheless, this traditional method could compromise data analysis accuracy due to human errors. The processing time to analyse the data would also be longer, especially for larger sample sizes. Therefore, this study proposed an automated negative lightning return strokes characterisation using a Brute-Force search algorithm, mainly on seven parameters, i.e. on zero-topeak rise time, 10-to-90% rise time, zero-crossing time, slow front time, slow front amplitude relative to peak, fast transition 10-to-90% rise time, and width dE/dt pulse at half peak value. For this purpose, the lightning-generated electric field measurement under different seasonal periods, was conducted at the College of Engineering, Universiti Tenaga Nasional (UNITEN), Selangor, Malaysia using a previously researched small-scale prallel-plate antenna. The criteria for each of the negative lightning return strokes parameters were established and defined based on specific mathematical equations. Accordingly, the proposed Brute-Force search algorithm characterised the negative lightning return strokes parameters based on the seven parameters. A total of 206 negative lightning return strokes waveforms were analysed and automatically characterised using the proposed algorithm. Comparisons of different data, including the manual data (i.e. obtained through the conventional method), data (automated) from a previous study, and the automated data (i.e. obtained using the proposed algorithm), were also carried out by evaluating the percentage difference, arithmetic mean, and standard deviation. The statistical analysis showed a good agreement between manual and automated data on each parameter, with a percentage difference observed between 0.08% and 6.88%. Bland-Altman plots analysis also showed that most of the plotted data were within the threshold limits. The results suggest that the proposed algorithm can provide an efficient structure and procedure by reducing the processing time, minimising human error, and eliminating the nonuniformity among users while characterising the lightning negative return strokes. Furthermore, the small-scale parallel-plate antenna that has been tested under real lightning condition can be operated as a portable, lightweight, and easy-to-install device for a lightning measurement system.