Pareto-optimal design of UHF antenna using modified non-dominated sorting genetic algorithm II

• Published in: IET microwaves, antennas & propagation Vol. 14; no. 12; pp. 1404 - 1410
• Main Authors: Bin, Feng, Wang, Feng, Chen, She, Sun, Qiuqin, Zhong, Lipeng, Lin, Shu
• Format: Journal Article
• Language: English
• Published: The Institution of Engineering and Technology 07.10.2020
• Subjects: adaptive crossover; broadband antennas; frequency 490.0 MHz to 1.52 GHz; genetic algorithms; high fidelity factor; labour division strategy; MNSGA‐II; modified nondominated sorting genetic algorithm II; multiobjective optimisation problems; mutation possibilities; Pareto optimisation; Pareto‐optimal design; Pareto‐optimal solutions; partial discharge detection; Research Article; UHF antenna; UHF antennas; ultra‐high‐frequency antenna; wideband antennas; UHF antenna; Pareto optimisation; labour division strategy; ultra-high-frequency antenna; wideband antennas; MNSGA-II; Pareto-optimal design; adaptive crossover; mutation possibilities; genetic algorithms; broadband antennas; high fidelity factor; Pareto-optimal solutions; modified nondominated sorting genetic algorithm II; UHF antennas; frequency 490.0 MHz to 1.52 GHz; multiobjective optimisation problems; partial discharge detection
• ISSN: 1751-8725; 1751-8733; 1751-8733
• DOI: 10.1049/iet-map.2020.0121

In this study, the ultra-high-frequency (UHF) antenna for partial discharge (PD) detection is optimised to simultaneously satisfy the requirements of low return loss and high fidelity factor (FF) in the frequency band of interest by using the modified non-dominated sorting genetic algorithm II (MNSGA-II). Based on the labour division strategy, the MNSGA-II adopts an adaptive crossover and mutation possibilities instead of the fixed ones, resulting in the significant improvement of convergence rate and exploration ability. One of the Pareto-optimal solutions is presented as the authors’ proposed UHF antenna, whose performance is compared with those of both the antenna optimised by genetic algorithm and the existing wideband antennas. The experimental results show that the proposed antenna with a compact size of 0.201 λL × 0.198 λL realises the reflection coefficient less than −10 dB from 490 MHz to 1.52 GHz, and its FFs in the face-to-face and side-by-side scenarios are 0.897 and 0.845, respectively. Furthermore, the simulation results reveal that the high FF of antenna greatly increases the accuracy of PD source localisation. It indicates that the MNSGA-II provides an excellent solution to the multiobjective optimisation problems in antenna design.