CAJ | 학술논문

针对粒子群优化算法容易早熟、收敛精度低等缺点,通过采用全变异策略、最大搜索速度自适应调整等策略得到了一种全变异粒子群优化算法,其中的全变异策略是在陷入早熟的条件下全体粒子参加变异,并且当把粒子看成染色体时,每一个基因等概率地参加变异,可以克服算法的早熟而继续优化,提高了算法的收敛精度.对Shubea函数进行实验的结果表明了算法的有效性.
침대입자군우화산법용역조숙、수렴정도저등결점,통과채용전변이책략、최대수색속도자괄응조정등책략득도료일충전변이입자군우화산법,기중적전변이책략시재함입조숙적조건하전체입자삼가변이,병차당파입자간성염색체시,매일개기인등개솔지삼가변이,가이극복산법적조숙이계속우화,제고료산법적수렴정도.대Shubea함수진행실험적결과표명료산법적유효성.
To overcome the premature and low convergence precision of particle swarm optimization,the whole Mutation Particle Swarm Optimization (MPSO) is proposed with whole mutation and the maximum velocity self-adjustment strategy,whole mutation strategy is adopted when PSO encounters premature,the particle is considered as chromosome and every gene has the same prob-ability to be mutated,the MPSO can overcome the local convergence of PSO and improves its convergence precision,the novel algorithm is used to solve the Shubert function optimization problem,the result shows that the algorithm is effective.