Self-adaptive differential evolution algorithm with improved mutation strategy

被引：46

作者：

Wang, Shihao ^{[1
,2
]}

Li, Yuzhen ^{[3
]}

Yang, Hongyu ^{[1
,2
]}

Liu, Hong ^{[2
]}

机构：

[1] Sichuan Univ, Sch Aeronaut & Astronaut, Chengdu 610065, Sichuan, Peoples R China

[2] Sichuan Univ, Natl Key Lab Air Traff Control Automat Syst Techn, Chengdu 610065, Sichuan, Peoples R China

[3] Shanghai Elect Apparat Res Inst, Shanghai 200063, Peoples R China

来源：

SOFT COMPUTING | 2018年 / 22卷 / 10期

基金：

中国国家自然科学基金;

关键词：

Differential evolution; Global optimization; Mutation strategy; Control parameters adaptation; Elite archive; GLOBAL OPTIMIZATION; PARTICLE SWARM; PARAMETERS; DESIGN;

D O I：

10.1007/s00500-017-2588-5

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Different mutation strategies and control parameters settings directly affect the performance of differential evolution (DE) algorithm. In this paper, a self-adaptive differential evolution algorithm with improved mutation strategy (IMSaDE) is proposed to improve optimization performance of DE. IMSaDE improves the "DE/rand/2" mutation strategy by incorporating elite archive strategy and control parameters adaptation strategy. Both strategies diversify the population and improve the convergence performance of the algorithm. IMSaDE was compared with eleven DE algorithms and six non-DE algorithms by using a set of 20 benchmark functions taken from the literature. Experimental results show that the overall performance of IMSaDE is better than the other competitors. In addition, the size of elite population has a significant impact on the performance of IMSaDE.

引用

页码：3433 / 3447

页数：15

共 33 条

[1]

[Anonymous], 2002, ADV INTELL SYST FUZZ

[2]

Babu BV, 2003, IEEE C EVOL COMPUTAT, P2696

[3] Population size reduction for the differential evolution algorithm [J].

Brest, Janez ;

Maucec, Mirjam Sepesy .

APPLIED INTELLIGENCE, 2008, 29 (03) :228-247

[4] Self-adapting control parameters in differential evolution: A comparative study on numerical benchmark problems [J].

Brest, Janez ;

Greiner, Saso ;

Boskovic, Borko ;

Mernik, Marjan ;

Zumer, Vijern .

IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION, 2006, 10 (06) :646-657

[5] A practical tutorial on the use of nonparametric statistical tests as a methodology for comparing evolutionary and swarm intelligence algorithms [J].

Derrac, Joaquin ;

Garcia, Salvador ;

Molina, Daniel ;

Herrera, Francisco .

SWARM AND EVOLUTIONARY COMPUTATION, 2011, 1 (01) :3-18

[6] Self-adaptive differential evolution algorithm with discrete mutation control parameters [J].

Fan, Qinqin ;

Yan, Xuefeng .

EXPERT SYSTEMS WITH APPLICATIONS, 2015, 42 (03) :1551-1572

[7] Global and local real-coded genetic algorithms based on parent-centric crossover operators [J].

Garcia-Martinez, C. ;

Lozano, M. ;

Herrera, F. ;

Molina, D. ;

Sanchez, A. M. .

EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2008, 185 (03) :1088-1113

[8] Self-adaptive differential evolution for feature selection in hyperspectral image data [J].

Ghosh, Ashish ;

Datta, Aloke ;

Ghosh, Susmita .

APPLIED SOFT COMPUTING, 2013, 13 (04) :1969-1977

[9]

Guo ZY, 2006, LECT NOTES COMPUT SC, V4221, P972

[10] Completely derandomized self-adaptation in evolution strategies [J].

Hansen, N ;

Ostermeier, A .

EVOLUTIONARY COMPUTATION, 2001, 9 (02) :159-195

← 1 2 3 4 →