一种改进的量子遗传算法研究

Improved Quantum Genetic Algorithm by Balancing Convergence and Diversity

针对量子遗传算法容易早熟收敛和陷入局部最优的问题,提出了一种均衡收敛性和多样性的改进量子遗传算法(IQGA),可结合小生境技术,通过使用惩罚函数修正个体的适应度来淘汰相似性高的个体,使种群具有良好的多样性,避免局部最优,同时,在进化过程中采用动态调整量子旋转角θ的更新策略,加快算法的收敛速度。实验结果表明,IQGA算法可以有效避免传统量子遗传算法早熟收敛,具有寻优能力强、收敛快等优点,适合于一般连续函数优化问题。

An improved quantum genetic algorithm by balancing convergence and diversity （IQGA） was proposed to overcome the problem of the premature and local optimization of the classic quantum genetic algorithm. In this algorithm, using niching can survive the fitter in similar individuals and modify the individuals by penalty function to enhance the diversity in population and avoid local optimization. What＇s more, adjusting quantum rotation angle dynamically can improve convergence speed. Simulations show that IQGA can avcid premature effectively and has good optimizing ability, good rate of convergence in 7,eneral continuous function optimization.