A Modified Differential Evolution Algorithm and Its Application to Engineering Problems

Logistic Map and Exponential Scaling Factor based Differential Evolution

International Journal of Swarm Intelligence Research ◽

10.4018/ijsir.2022010119 ◽

2022 ◽

Vol 13 (1) ◽

pp. 0-0

Keyword(s):

Differential Evolution ◽

Logistic Map ◽

Differential Evolution Algorithm ◽

Scaling Factor ◽

Exploration And Exploitation ◽

Engineering Problems ◽

Evolutionary Technique ◽

Evolution Algorithm ◽

State Of Art ◽

Modified Algorithm

Differential evolution (DE), an important evolutionary technique, enhances its parameters such as, initialization of population, mutation, crossover etc. to resolve realistic optimization issues. This work represents a modified differential evolution algorithm by using the idea of exponential scale factor and logistic map in order to address the slow convergence rate, and to keep a very good equilibrium linking exploration and exploitation. Modification is done in two ways: (i) Initialization of population and (ii) Scaling factor.The proposed algorithm is validated with the aid of a 13 different benchmark functions taking from the literature, also the outcomes are compared along with 7 different popular state of art algorithms. Further, performance of the modified algorithm is simulated on 3 realistic engineering problems. Also compared with 8 recent optimizer techniques. Again from number of function evaluations it is clear that the proposed algorithm converses more quickly than the other existing algorithms.

Download Full-text

Enhanced Directed Differential Evolution Algorithm for Solving Constrained Engineering Optimization Problems

International Journal of Applied Metaheuristic Computing ◽

10.4018/ijamc.2019010101 ◽

2019 ◽

Vol 10 (1) ◽

pp. 1-28 ◽

Cited By ~ 8

Author(s):

Ali Wagdy Mohamed ◽

Ali Khater Mohamed ◽

Ehab Z. Elfeky ◽

Mohamed Saleh

Keyword(s):

Differential Evolution ◽

Optimization Problems ◽

Differential Evolution Algorithm ◽

Test Problems ◽

New Mutation ◽

De Algorithm ◽

Engineering Problems ◽

Evolution Algorithm ◽

Mutation Scheme ◽

Exploration Exploitation

The performance of Differential Evolution is significantly affected by the mutation scheme, which attracts many researchers to develop and enhance the mutation scheme in DE. In this article, the authors introduce an enhanced DE algorithm (EDDE) that utilizes the information given by good individuals and bad individuals in the population. The new mutation scheme maintains effectively the exploration/exploitation balance. Numerical experiments are conducted on 24 test problems presented in CEC'2006, and five constrained engineering problems from the literature for verifying and analyzing the performance of EDDE. The presented algorithm showed competitiveness in some cases and superiority in other cases in terms of robustness, efficiency and quality the of the results.

Download Full-text

Logistic Map and Exponential Scaling Factor-Based Differential Evolution

International Journal of Swarm Intelligence Research ◽

10.4018/ijsir.2022010108 ◽

2022 ◽

Vol 13 (1) ◽

pp. 1-15

Author(s):

Katyayani Kashyap ◽

Sunil Pathak ◽

Narendra Singh Yadav

Keyword(s):

Differential Evolution ◽

Logistic Map ◽

Differential Evolution Algorithm ◽

Scaling Factor ◽

Exploration And Exploitation ◽

Engineering Problems ◽

Evolutionary Technique ◽

Evolution Algorithm ◽

State Of Art ◽

Modified Algorithm

Differential evolution (DE), an important evolutionary technique, enhances its parameters such as, initialization of population, mutation, crossover etc. to resolve realistic optimization issues. This work represents a modified differential evolution algorithm by using the idea of exponential scale factor and logistic map in order to address the slow convergence rate, and to keep a very good equilibrium linking exploration and exploitation. Modification is done in two ways: (i) Initialization of population and (ii) Scaling factor.The proposed algorithm is validated with the aid of a 13 different benchmark functions taking from the literature, also the outcomes are compared along with 7 different popular state of art algorithms. Further, performance of the modified algorithm is simulated on 3 realistic engineering problems. Also compared with 8 recent optimizer techniques. Again from number of function evaluations it is clear that the proposed algorithm converses more quickly than the other existing algorithms.

Download Full-text