發(fā)布時(shí)間：2018-05-31 01:09

  本文選題：光學(xué)優(yōu)化算法 + 適應(yīng)度　； 參考：《計(jì)算機(jī)工程與應(yīng)用》2017年12期

【摘要】：光學(xué)優(yōu)化算法是一種新型優(yōu)化算法,源自物理學(xué)中的光學(xué)原理。針對(duì)基本光學(xué)優(yōu)化算法中適應(yīng)度函數(shù)隨進(jìn)化過程恒定不變導(dǎo)致算法搜索能力差、精度低等不足之處,結(jié)合遺傳算法中自適應(yīng)度的改進(jìn)方法,提出一種可隨進(jìn)化代數(shù)動(dòng)態(tài)調(diào)整的非線性適應(yīng)度函數(shù),改進(jìn)了光學(xué)優(yōu)化算法的適應(yīng)度函數(shù)。通過一系列典型的基準(zhǔn)函數(shù)測(cè)試了改進(jìn)算法的性能,實(shí)驗(yàn)結(jié)果驗(yàn)證了改進(jìn)算法的可行性與有效性。
[Abstract]:Optical optimization algorithm is a new optimization algorithm derived from the optical principles in physics. The fitness function of the basic optical optimization algorithm is invariant with the evolution process, which leads to the poor searching ability and low precision of the algorithm. The improved method of adaptive degree is combined with the genetic algorithm. A nonlinear fitness function, which can be dynamically adjusted with evolution algebra, is proposed, and the fitness function of optical optimization algorithm is improved. The performance of the improved algorithm is tested by a series of typical benchmark functions. The experimental results show that the improved algorithm is feasible and effective.
【作者單位】： 上海理工大學(xué)管理學(xué)院;
【基金】：國家自然科學(xué)基金(No.71401106) 上海市高峰高原學(xué)科建設(shè)項(xiàng)目
【分類號(hào)】：O43;TP18
，

本文編號(hào)：1957638