A novel niche genetic algorithm with local search ability
Junhua Gu, Na-Na Li, Qing Tan, Wei Wei · 2007
The insufficiency of local search and slow convergence in later generations are two main disadvantages of niche genetic algorithm (NGA). In this paper, we propose an improved novel niche genetic algorithm with local search ability. Depending on the number of iteration, the new algorithm adopts the mechanism of crossover operator and mutation operator in niche population instead of between different niches to make the searching more effective. This new method is used in Shubert function optimization and experimental results show its superiority compared with GA and NGA.