Dynamic modeling of flashover of polymer insulators under polluted conditions based on HGA-PSO algorithm

• Published in: Electric power systems research 2022-04, Vol.205, p.107728, Article 107728
• Main Authors: Fahimi, Navid, Sezavar, Hamid Reza, Shayegani Akmal, Amir Abbas
• Format: Article
• Language: English
• Subjects: Algorithms; Combinations (mathematics); Correlation analysis; Dynamic modeling; Dynamic models; Electric potential; Electricity; Flashover; Flow charts; Insulation; Insulators; Leakage current; Mathematical models; Online monitoring; Particle swarm optimization; Pollution flashover; Pollution monitoring; Polymer insulator; Polymers; Voltage
• ISSN: 0378-7796; 1873-2046
• DOI: 10.1016/j.epsr.2021.107728

•Dynamic modeling approach is utilized to understand pollution flashover of insulators.•Artificial intelligence is used to develop dynamic model of pollution flashover of insulators.•There are close correlations between experimental data and dynamic model of pollution flashover.•Development of an online monitoring system through dynamic modeling of pollution flashover of insulators is an effective tool to predict flashover. This paper investigates the pre-flashover conditions of polymer insulators under polluted condition. This investigation is done by combination of analytical mathematical formulations, which were deduced in previous researches with AI based on HGA-PSO. The aim of this novel approach is to find proper correlation between experimental data of voltage and leakage current of insulators to theoretical formulations. Correlations are obtained and compared between discharge resistance and leakage currents for different cases of the pre-flashover conditions. Some experiments on real medium voltage polymer insulators under polluted conditions are done and real-time data of voltage and leakage currents are recorded as the voltage is increased step-by-step until flashover occurs. The experimental data are used by the proposed approach to better understand the pollution flashover behavior. Intelligent approach can be traced through a proposed flowchart which illustrates calculation procedure from the beginning point to the final point of identification of the pre-flashover conditions. Results of the intelligent algorithm show proper correspondence to the analytical formulations. In addition, the proposed algorithm is able to predict pre-flashover conditions properly using ANN classification. The proposed procedure can be an effective tool to develop an online monitoring system in order to prevent pollution flashover of polymer insulators.