安徽大学学报(自然科学版)
安徽大學學報(自然科學版)
안휘대학학보(자연과학판)
JOURNAL OF ANHUI UNIVERSITY(NATURAL SCIENCES EDITION)
2014年
3期
16-23
,共8页
汪继文%杨丹%邱剑锋%王心灵
汪繼文%楊丹%邱劍鋒%王心靈
왕계문%양단%구검봉%왕심령
群智能%非线性方程组%人工蜂群算法%差分进化%随机向量
群智能%非線性方程組%人工蜂群算法%差分進化%隨機嚮量
군지능%비선성방정조%인공봉군산법%차분진화%수궤향량
swarm intelligence%nonlinear equations%artificial bee colony algorithm%differential evolution%random vector
针对传统的人工蜂群算法在处理单峰问题时收敛速度较慢、多峰时易陷入局部最优等缺点,通过借鉴差分进化算法中变异算子的作用,提出了一种改进的人工蜂群算法。该改进算法在对蜜源邻域的搜索过程中引入了个体当前最优值及随机向量,从而加快算法的收敛速度,并且在一定程度上防止多峰问题易陷入局部最优的不足,提高算法的搜索能力。最后将改进的算法应用到求解基本函数和非线性方程组上,测试改进算法的性能。结果表明,改进的算法能够有效避免陷入局部最优,并能较大幅度地提高收敛速度和收敛精度。
針對傳統的人工蜂群算法在處理單峰問題時收斂速度較慢、多峰時易陷入跼部最優等缺點,通過藉鑒差分進化算法中變異算子的作用,提齣瞭一種改進的人工蜂群算法。該改進算法在對蜜源鄰域的搜索過程中引入瞭箇體噹前最優值及隨機嚮量,從而加快算法的收斂速度,併且在一定程度上防止多峰問題易陷入跼部最優的不足,提高算法的搜索能力。最後將改進的算法應用到求解基本函數和非線性方程組上,測試改進算法的性能。結果錶明,改進的算法能夠有效避免陷入跼部最優,併能較大幅度地提高收斂速度和收斂精度。
침대전통적인공봉군산법재처리단봉문제시수렴속도교만、다봉시역함입국부최우등결점,통과차감차분진화산법중변이산자적작용,제출료일충개진적인공봉군산법。해개진산법재대밀원린역적수색과정중인입료개체당전최우치급수궤향량,종이가쾌산법적수렴속도,병차재일정정도상방지다봉문제역함입국부최우적불족,제고산법적수색능력。최후장개진적산법응용도구해기본함수화비선성방정조상,측시개진산법적성능。결과표명,개진적산법능구유효피면함입국부최우,병능교대폭도지제고수렴속도화수렴정도。
The traditional artificial bee colony algorithm converged slowly in dealing with the unimodal problems and got into local optimum easily when dealing with the multimodal problems. So a modified artificial bee colony algorithm was proposed, which took examples from the differential evolution operators. This modified algorithm introduced individual current optimal value and random vector in the random search process on the neighborhood of nectar, which sped up the convergence rate of the unimodal problems and to some level, prevented multimodal problems from easily getting into local optimum. This improved the search ability of the algorithm. Finally, the improved algorithm was used to solve some basic functions and nonlinear equations to test the performance of it. Experimental results showed that the new algorithm not only avoided effectively being trapped in local minima but also outperformed others in terms of convergence rate and accuracy.