An Improved Genetic Algorithm and Its Application in Global Optimization of Electromagnetic Field Problems

陈旭东,乔静秋,熊素铭,倪光正,杨仕友
DOI: https://doi.org/10.3969/j.issn.1007-449x.2001.04.005
2001-01-01
Abstract:An improved genetic algorithm for global optimization of electromagnetic field roblems isdeveloped in this paper. The reason for the slow convergence of the standard genetic algorithm is analyzed at first, then the improved genetic algorithm is proposed to expeditethe convergence. Techniques as partition of optimization process(random search stage and local optimization stage), novel criterion for new generation creation, adaptive updating of mutation probability and new stop criterion are introduced in the improved algorithm. The numeral results on a mathematical test and a practical problem both show the proposed method not only obtains better results, but also is timesaving and more robust compared with the standard genetic algorithm.
What problem does this paper attempt to address?