摘要
粒子群算法是一种全局智能优化算法,针对该算法在早期迭代中容易造成局部极值,在后期迭代中容易造成种群的多样性消失,使得算法收敛速度减慢,求解质量不高等缺点。本文提出通过收敛吸引因子粒子来获得局部最优值;加入扰动函数来更新粒子的速度来提高了算法整体效率。经典测试函数证明本文算法性能明显优于基本PSO算法,同时在算法复杂度方面优于其他的智能算法,有效地提升了算法的求解精度。
Particle swarm optimization is a global intelligent improvement optimization. Aiming at itsdefects such as being easy to cause local extremum at the earlier stage but the disappearance of diversityduring the later iteration, thus convergence speed gets slow down with solutions of poor quality,convergence attraction factors are introduced in this paper to get the local optimal value and disturbancefunction is added to update the speed of particles and improve overall efficiency of the algorithm.Classical test function shows that algorithm in this paper has obviously superior performance than thebasic PSO algorithm. Meanwhile, it is superior to other intelligent algorithms in terms of complexity,effectively improving precision of the algorithm’s solutions.
出处
《科技通报》
北大核心
2016年第8期154-159,共6页
Bulletin of Science and Technology
关键词
粒子群算法
收敛吸引因子
扰动函数
particle swarm optimization
convergence attraction factor
disturbance function