A Customized Differential Evolutionary Algorithm for Bounded Constrained Optimization Problems

Mashwani, Wali Khan; Rehman, Zia Ur; Bakar, Maharani A.; Koçak, Ismail; Fayaz, Muhammad

doi:https://doi.org/10.1155/2021/5515701

Complexity

On this page

Abstract Introduction Discussion Conclusion Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Research Article | Open Access

Volume 2021 | Article ID 5515701 | https://doi.org/10.1155/2021/5515701

A Customized Differential Evolutionary Algorithm for Bounded Constrained Optimization Problems

Wali Khan Mashwani,¹Zia Ur Rehman,¹Maharani A. Bakar,²Ismail Koçak,³and Muhammad Fayaz⁴

Academic Editor: Atif Khan

Received16 Jan 2021

Revised16 Feb 2021

Accepted24 Feb 2021

Published10 Mar 2021

Abstract

Bound-constrained optimization has wide applications in science and engineering. In the last two decades, various evolutionary algorithms (EAs) were developed under the umbrella of evolutionary computation for solving various bound-constrained benchmark functions and various real-world problems. In general, the developed evolutionary algorithms (EAs) belong to nature-inspired algorithms (NIAs) and swarm intelligence (SI) paradigms. Differential evolutionary algorithm is one of the most popular and well-known EAs and has secured top ranks in most of the EA competitions in the special session of the IEEE Congress on Evolutionary Computation. In this paper, a customized differential evolutionary algorithm is suggested and applied on twenty-nine large-scale bound-constrained benchmark functions. The suggested C-DE algorithm has obtained promising numerical results in its 51 independent runs of simulations. Most of the 2013 IEEE-CEC benchmark functions are tackled efficiently in terms of proximity and diversity.

1. Introduction

Bound optimization is the mathematical process of optimizing an objective function in the presence of constraints imposed on the decision space. These variables in the decision space may be continuous, discrete, or mixed. The basic elements of optimization are decision variables, objective function, and constrained functions. The standard form for a single-objective, nonlinear, constrained optimization problem is described as follows [1]: where is the candidate solution with decision variables/real parameters, is the objective function, represents inequality constrained functions, denotes equality constrained functions, and are bound constrained over the search space . An optimum solution that is optimal in some certain neighbourhood is called a local optimum. Mathematically, by a local optimum, we mean a solution such that for all in some neighbourhood of , Then, optimum solution is called global optimum solution if .

In the last two decades, various optimization methods were developed and are still developing for dealing with a diverse test suite of benchmark functions and real-world optimization problems [2–6]. These algorithms are inspired by Darwin’s theory of evolution with Darwinian’s principle “Survival of the fittest”. The main intrinsic operators in the framework of EAs to evolve population include mutation and crossover operators, by using mutation and crossover operator with elitism scheme [7]. EAs are nature-inspired-based techniques [8–10]. In general, EAs can be categorized into nine different groups including the biology-based [11, 12], physics-based [13, 14], social-based [15–18], music-based [19], chemical-based [20], sport-based [21], mathematics-based [22], swarm-based [23–29], and hybrid methods [30–37]. All these algorithms operate on a uniform and random set of solutions generated within the search domain of the given optimization and search problems. Among the evolutionary algorithms, the genetic algorithm (GA) [38], evolutionary strategy (ES) [39], evolutionary programming (EP) [5, 29], and genetic programming (GP) [40] are classical paradigms in evolutionary computing field [3]. Differential evolution (DE) [41], particle swarm optimization (PSO) [24], and ant colony optimization (ACO) [23, 42, 43], cuckoo search (CS) [44], and teaching learning-based optimization (TLBO) algorithm [45, 46] are comparatively recently developed EAs and successfully applied to various test suites of optimization problems and many real-world problems [18, 47–50].

Differential evolution (DE) is one of the well-established population-based evolutionary algorithms, which was first introduced by Storn and Price for solving the Chebyshev polynomial fitting problems [51, 52]. DE is capable of handling nonlinear, linear, and multimodel objective functions and solving directional preferences over existing relays in the power system. The DE has also worked to train the weight of the neural network to deal with real-world problems. The basic idea of this technique is to avoid repetition of the existing set of solutions. Classic DE has two main control parameters and that need to be settled efficiently. The settlement of these parameters including the scaling factor, , and the probability of crossover,, is as discussed in [31, 41, 53–61]. Crossover is the main source of exploration, the mutation is utilized for exploitation purposes, and selection operators brought down the pressure for the survival of fittest individuals to evolve the population. The adjustment of various parameters used in search operator implementation is mainly problem-dependent, and their proper settings are quite difficult and time-consuming while performing trial and error experiments [62].

Different search operators behave differently at different levels of the optimization search process while dealing with complicated optimization and search problems. It is quite difficult to determine that this particular crossover or mutation or selection operator is useful in this particular proposition optimization process. In this paper, a customized differential evolutionary algorithm is suggested for solving benchmark functions designed for the special session of the 2013 IEEE Congress of Evolutionary Computation (CEC’13) [63]. The proposed C-DE algorithm employs more than one mutation operator flowing by crossover to generate a new set of solutions for the next generation of the evolutionary search process. Experimental results demonstrated that the proposed C-DE algorithm is more promising as compared to the algorithm that employs a single mutation operator for population evolution.

The rest of the paper is organized as follows. Section 2 introduces the framework of the proposed customized differential evolutionary algorithm. Section 3 demonstrates the experimental results and characteristics of the used benchmark functions. Section 4 finally concludes this paper.

2. A Customized Differential Evolutionary Algorithm for Numerical Optimization Problems

Differential evolution (DE) is a stochastic algorithm that was first developed by Kenneth Price in 1994 [40]. DE is one of the most promising global search techniques that converge toward the known optimal solution of the problem at hand in a significant manner [60]. In general, DE has five different mutation strategies; among them, rand/1 and best/1rand/1 have performed better in our case. Among the five DE variants, DE rand/1 has faced premature convergence issues and slow convergence behaviour as compared to DE best/1rand/1. The proposed customized DE employs DE rand/1 and DE best/1rand/1 mutation strategies simultaneously to alleviate premature convergence of original differential evolution. The customized DE does the switching of aforementioned mutation strategies with linearly reducing weight parameter from 0.25 to 0.01 during their whole course of optimization. (Figure 1).

(1)	Set Pop , Max Ftn Evaluations = , lower scaling bound = , upper scaling bound = , crossover probability = ;
(2)	Randomly generate initial of size ;
(3)	Set Function Evaluations Counter: ;
(4)	Evaluate for each ;
(5)	Set ;
(6)	such that ;
(7)	whiledo
(8)	Generate from using rand/1 mutation strategy;
(9)	Use binomial crossover to generate ;
(10)	Evaluate for each ;
(11)	Set ;
(12)	Update by comparing with through their fitness values;
(13)	such that ;
(14)	end while

(1)	Set Pop Size = , Max Ftn Evaluations = , lower scaling bound = , upper scaling bound = , crossover probability = ;
(2)	Randomly generate initial of size ;
(3)	Set Function Evaluations Counter: ;
(4)	Evaluate for each ;
(5)	Set ;
(6)	such that ;
(7)	whiledo
(8)	Generate from using rand/2 mutation strategy;
(9)	Use binomial crossover to generate ;
(10)	Evaluate for each ;
(11)	Set ;
(12)	Update by comparing with through their fitness values;
(13)	such that ;
(14)	end while

(1)	Set Pop Size = , Max Ftn Evaluations = , lower scaling bound = , upper scaling bound = , crossover probability = ;
(2)	Randomly generate initial of size ;
(3)	Set Function Evaluations Counter: ;
(4)	Evaluate for each ;
(5)	Set ;
(6)	such that ;
(7)	whiledo
(8)	Generate from using best/rand/2 strategy;
(9)	Use binomial crossover to generate ;
(10)	Evaluate for each ;
(11)	Set ;
(12)	Update by comparing with through their fitness values;
(13)	such that ;
(14)	end while

(1)	Set Pop Size = , Max Ftn Evaluations = , lower scaling bound = , upper scaling bound = , crossover probability = ;
(2)	Randomly generate initial of size ;
(3)	Set Function Evaluations Counter: ;
(4)	Evaluate for each ;
(5)	Set ;
(6)	such that ;
(7)	whiledo
(8)	Generate from using current/rand/2 mutation strategy;
(9)	Use binomial crossover to generate ;
(10)	Evaluate for each ;
(11)	Set ;
(12)	Update by comparing with through their fitness values;
(13)	such that ;
(14)	end while

(1)	Set Pop Size = , Max Ftn Evaluations = , lower scaling bound = , upper scaling bound = , crossover probability = ;
(2)	Randomly generate initial of size ;
(3)	Set Function Evaluations Counter: ;
(4)	Evaluate for each ;
(5)	Set ;
(6)	such that ;
(7)	whiledo
(8)	Generate from using best/1rand/1 strategy;
(9)	Use binomial crossover to generate ;
(10)	Evaluate for each ;
(11)	Set ;
(12)	Update by comparing with through their fitness values;
(13)	such that ;
(14)	end while

Mutation strategies are the most distinctive components of differential evolution which are described as follows [64]:(1)Rand/1: mathematically, it is formulated as follows: where , , and are randomly taken distinct solutions from population other than (target solution). is the scaling factor which is randomly generated. is the mutant vector. In Algorithm 1, the mutation strategy as described in equation (2) is used and solved the CEC-2013 unconstrained suite of benchmark functions. The abovementioned mutation strategy is good in exploration while comparatively weak in exploitation because of which it does not converge efficiently keeping in view the average value of most of the used benchmark functions (Algorithm 2).(2)Best/rand/1: mathematically, it is formulated as follows: where is the best solution in the current population, and are randomly chosen distinct solutions, and is the target solution. is the scaling factor which is randomly generated. is the mutant vector. In Algorithm 3, the mutation strategy as described in equation (3) has shown better performance in exploitation and comparatively not good in exploration that induces a premature convergence.(3)Current/rand/2: mathematically, it is formulated as follows: where is the target solution, , , , and are randomly taken distinct solutions from population other than target solution. is the scaling factor which is randomly generated. is the mutant vector. The mutation strategy as described in equation (4) is employed in Algorithm 4 for creating a new solution. The above-described mutation strategy is good in exploitation while comparatively weak in exploration.(4)Rand/2: mathematically, it is formulated as follows: where , , , , and are randomly taken distinct solutions from population other than target solution. is the scaling factor which is randomly generated. is the mutant vector. The mutation strategy as described in equation (5) has been used in Algorithm 4 and performed good in exploration while less effective in exploitation.(5)Best/rand/2: mathematically, it is formulated as follows: where is the best solution in the current population and , , , and are taken randomly. is the scaling factor which is randomly generated within the specified bound. is the mutant vector. The mutation strategy as described in equation (6) has been employed in Algorithm 4 to maintain exploitation versus exploration issue in order to avoid premature convergence during population evolution.(6)Best/rand/1: mathematically, it is formulated as follows: where , , , , and are chosen randomly. The scaling factor, , is adjusted in all algorithms, and is the mutant vector as used in Algorithm 5.

The proposed C-DE technique as outlined in Algorithm 6 employs the mutation strategies as described in equations (2) and (6) along with crossover as described in equation (8) to create its new set of solutions for the next generation of population evolution.

(1)	Set Pop Size = , Max Ftn Evaluations = , lower scaling bound = ,upper scaling bound = ,crossover probability = ;
(2)	Randomly generate initial of size ;
(3)	Set Function Evaluations Counter: ;
(4)	Evaluate for each ;
(5)	Set ;
(6)	such that ;
(7)	whiledo
(8)	Take randomly generated real number between 0 and 1;
(9)	ifthen
(10)	Generate from using rand/1 strategy;
(11)	else
(12)	Generate from usingbest1/rand1 strategy;
(13)	end if
(14)	Use binomial crossover to generate ;
(15)	Evaluate for each ;
(16)	Set ;
(17)	Update by comparing and through their fitness values;
(18)	such that ;
(19)	end while

2.1. Crossover

In the current research, a binomial crossover operator is used. In the binomial crossover, every target vector is mixed with its corresponding mutant vector to get the trial vector as follows:where is the coordinate counter, is the mutant vector, is the probability of crossover, is a randomly generated integer, denotes the dimension of search space, is the target vector, and is the trial vector corresponding to .

2.2. Selection

The selection scheme for parent solution and their trial solution vector compete with each other based on their corresponding fitness values. The selection scheme is illustrated as follows:where is the trial vector, is the target vector, and and are the fitness values of the -trial and -target vector, respectively.

3. Simulation Results, Comparison, and Discussion

Due to the flurry of EAs recently developed, their performances are mainly analyzed by using different test suites of optimization and search problems. In the last few years, several unconstrained benchmark functions (i.e., bound-constrained) and constrained test problems are designed in special sessions of the IEEE Conference on Evolutionary Computation (IEEE-CEC) series for EA competition [63]. In a study of this paper, we have used twenty-eight different unconstrained continuous benchmark functions whose characteristics are given in Table 1. The first five test problems, , are unimodal functions. Most of the evolutionary algorithms are getting trap in the local basin of attraction while solving these test problems. The problems from to are multimodal functions containing more than one local minima. These problems are much difficult to deal with as compared to unimodal functions. The functions from are composite functions.

3.1. PC Configuration and Simulation Environment

The simulations were conducted on a PC with 8 GB RAM and Intel(R)Core(TM)i3-7100U CPU @ 2.40 GHz processor. All simulations are done in MATLAB R2013a environment with . Each benchmark function has been solved in different dimensions including by executing 51 times independently. Maximum function evaluation was fixed to . The algorithms for each problem are run 51 times independently to obtain the final best solution. denotes the switching of mutation strategies.

Tables 2 and 3 present the numerical results of C-DE as outlined in Algorithm 6 in comparison with DE-1 as described in Algorithm 1 and DE-5 hereby outlined in Algorithm 5. The numerical results for to benchmark functions in terms of minimum, maximum, average, median, and standard deviation in function values are summarized in Table 2. Similarly, the numerical results for the benchmark functions denoted by to are given in Table 3.

Tables 4 and 5 demonstrate the numerical results in respect of the proposed C-DE as outlined in Algorithm 6 versus DE-1 as given in Algorithm 1 and DE-5 hereby outlined in Algorithm 5. The numerical results for to benchmark functions in terms of minimum, maximum, average, median, and standard deviation in function values are summarized in Table 4. Similarly, the numerical results regarding the benchmark functions denoted by to are given in Table 5.

Tables 6 and 7 provide the experimental results approximated by the proposed C-DE herewith outlined in Algorithm 6 against DE-1 as given in Algorithm 1 and DE-5 hereby outlined in Algorithm 5. The numerical results for to benchmark functions in terms of minimum, maximum, average, median, and standard deviation in function values are summarized in Table 6. Similarly, the numerical results regarding the benchmark functions denoted by to are given in Table 7.

Figure 2 displays the convergence graph of the to benchmark functions solved by the proposed C-DE algorithm versus DE-1 and DE-5 after 51 independent runs of simulation with different random seeds. The first panel of Figure 2 depicts the to benchmark functions as solved in ten dimensions, the second panel is for thirty dimensions, and the third panel is for fifty dimensions.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

Figure 3 displays the convergence graph of the to benchmark functions solved by the proposed C-DE algorithm versus DE-1 and DE-5 after 51 independent runs of simulation with different random seeds. The first panel of Figure 3 depicts the to benchmark functions as solved in ten dimensions, the second panel is for thirty dimensions, and the third panel is for fifty dimensions.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

Figure 4 displays the convergence graph of the to benchmark functions solved by the proposed C-DE algorithm versus DE-1 and DE-5 after 51 independent runs of simulation with different random seeds. The first panel of Figure 4 depicts the to benchmark functions as solved in ten dimensions, the second panel is for thirty dimensions, and the third panel is for fifty dimensions.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

Figure 5 displays the convergence graph of the to benchmark functions solved by the proposed C-DE algorithm versus DE-1 and DE-5 after 51 independent runs of simulation with different random seeds. The first panel of Figure 5 depicts the to benchmark functions as solved in ten dimensions, the second panel is for thirty dimensions, and the third panel is for fifty dimensions.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

Figure 6 displays the convergence graph of the to benchmark functions solved by the proposed C-DE algorithm versus DE-1 and DE-5 after 51 independent runs of simulation with different random seeds. The first panel of Figure 6 depicts the to benchmark functions as solved in ten dimensions, the second panel is for thirty dimensions, and the third panel is for fifty dimensions.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)

(n)

(o)

Figure 7 displays the convergence graph of the to benchmark functions solved by the proposed C-DE algorithm versus DE-1 and DE-5 after 51 independent runs of simulation with different random seeds. The first panel of Figure 7 depicts the to benchmark functions as solved in ten dimensions, the second panel is for thirty dimensions, and the third panel is for fifty dimensions.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

4. Conclusion

In this paper, three variants of differential evolutionary algorithms, namely, DE-1, DE-5, and C-DE, are suggested and their numerical results in terms of , , , , and are obtained for the unconstrained benchmark functions designed for the special session on evolutionary algorithm competition (the 2013 IEEE Conference on Evolutionary Computation). Among the three DE variants, C-DE as outlined in Algorithm 6 has obtained better experimental results as compared to DE-1 as reported in Algorithm 1 and DE-5 described in Algorithm 5. Based on comparative analysis, we conclude the following: The fitness values of C-DE are much better for mostly test problems with subjects to the “No Free Lunch Theorem” concept in the parlance of evolutionary computing communities. Based on the objective function values, C-DE has again prominent position among the other two compared algorithms in order to solve most benchmark functions with dimensions keeping in view the criterion of “No Free Lunch Theorem” which guarantees the transparency of the manuscript. The objective function values indicate in respect of C-DE much better than the average function values of other DE variant algorithms. The objective function values found by C-DE are much promising keeping in view the median values of the competing algorithms which reflect the fairness of the proposed algorithm. Competitors play a comparable role keeping in view the standard deviation values obtained for the used benchmark function. The efficiency and robustness of the proposed algorithm are satisfactory which can be observed from the error function values as plotted against the number of iterations.

In future, we will be planning for the following tasks: To check the efficiency of C-DE on other existing unconstrained suites to more elaborate the importance of the proposed algorithm To do parameter sensitivity to ensure the correct use of parameters on specific type problems is included in the future work For checking the robustness and efficiency of the proposed algorithm, we are intended to compare it with the advanced version of DE such as JADE, SHADE, and IDE For customization, the proposed algorithm with other advanced variants of DE can be come in count in future

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request. Furthermore, the MATLAB code of the benchmark functions is available on https://www3.ntu.edu.sg/home/epnsugan/

Conflicts of Interest

The authors declare that they have no conflicts of interest to report regarding the present study.

Acknowledgments

The authors acknowledge the support of the Research Intensified Grant Scheme (RIGS) Fasa 1/2019, Vot Number 55192/7, Universiti Malaysia Terengganu.

References

G. Wu, R. Mallipeddi, and P. Suganthan, Problem Definitions and Evaluation Criteria for the CEC 2017 Competition on Constrained Real-Parameter Optimization, National University of Defense Technology, Changsha, Hunan, 2017.
I. N. daSilva, D. H. Spatti, R. A. Flauzino, L. H. B. Liboni, and S. F. dos ReisAlves, Artificial Neural Networks: A Practical Course, Springer Publishing Company, Berlin, Germany, 1st edition, 2016.
A. E. Eiben and J. E. Smith, Introduction to Evolutionary Computing, Springer Publishing Company, Berlin, Germany, 2nd edition, 2015.
S.-C. Fang and S. Puthenpura, Linear Optimization and Extensions: Theory and Algorithms, Prentice-Hall, Inc., Upper Saddle River, NJ, USA, 1993.
J. R. Koza, Genetic Programming: On the Programming of Computers by Means of Natural Selection, Springer, Berlin, Germany, 1992.
R. E. Miller, Optimization: Foundations and Applications, John Wiley & Sons, Hoboken, NJ, USA, 1999.
X. Yu and M. Gen, “Introduction to evolutionary algorithms,” in Decision Engineering, R. Roy, Ed., Springer, New York, NY, USA, 2011.
View at: Google Scholar
T. Back, Evolutionary Algorithms in Theory and Practice: Evolution Strategies, Evolutionary Programming, Genetic Algorithms, Oxford University Press, Inc., New York, NY, USA, 1996.
E. E. Agoston, Introduction to Evolutionary Computing, Springer Publishing Company, Berlin, Germany, 2nd edition, 2015.
R. Mallipeddi and P. N. Suganthan, “Differential evolution algorithm with ensemble of populations for global numerical optimization,” Opsearch, vol. 46, no. 2, pp. 184–213, 2009.
View at: Publisher Site | Google Scholar
S. Akyol and B. Alatas, “Plant intelligence based metaheuristic optimization algorithms,” Artificial Intelligence Review, vol. 47, no. 4, pp. 417–462, 2017.
View at: Publisher Site | Google Scholar
B. I. S. MuhammadSulaiman, S. Abdellah, and O. B. Kirikchi, “A plant propagation algorithm for constrained engineering optimisation problems,” Mathematical Problems in Engineering, vol. 2014, p. 10, 2014.
View at: Google Scholar
S. Hong, D. Han, K. Cho, J. S. Shin, and J. Noh, “Physics-based full-body soccer motion control for dribbling and shooting,” ACM Transactions on Graphics, vol. 38, no. 4, pp. 81–74, 2019.
View at: Google Scholar
N. Siddique and H. Adeli, “Physics-based search and optimization: inspirations from nature,” Expert Systems, vol. 33, no. 6, pp. 607–623, 2016.
View at: Publisher Site | Google Scholar
F. Johari, A. Zain, N. Mustaffa, and A. Udin, “Firefly algorithm for optimization problem,” Applied Mechanics and Materials, vol. 421, 2013.
View at: Google Scholar
L. Li and F. Liu, Group Search Optimization for Applications in Structural Design, Springer, Berlin, Germany, 2011.
X.-S. Yang, “A new metaheuristic bat-inspired algorithm,” in Nature inspired cooperative strategies for optimization (NICSO 2010), pp. 65–74, Springer, Granada, Spain, May 2010.
View at: Google Scholar
X. Yu and M. Gen, Introduction to Evolutionary Algorithms, Springer Science & Business Media, Berlin, Germany, 2010.
J. H. Jeong and C. W. Ahn, “Automatic evolutionary music composition based on multi-objective genetic algorithm,” in 18th Asia Pacific Symposium on Intelligent and Evolutionary Systems, Handa, Ed. Ishibuchi, Y.-S. Ong, and K.-C. Tan, Eds., Springer International Publishing, Cham, Switzerland, 2015.
View at: Google Scholar
F. T. Silva, M. X. Silva, and J. C. Belchior, “A new genetic algorithm approach applied to atomic and molecular cluster studies,” Frontiers in Chemistry, vol. 7, p. 707, 2019.
View at: Publisher Site | Google Scholar
L. While and G. Kendall, “Scheduling the English football league with a multi-objective evolutionary algorithm,” in Parallel Problem Solving from Nature–PPSN XIII, Springer International Publishing, Cham, Switzerland, 2014.
View at: Google Scholar
H. Mühlenbein and T. Mahnig, Mathematical Analysis of Evolutionary Algorithms, Springer US, Boston, MA, USA, 2002.
C. Blum, “Ant colony optimization: introduction and recent trends,” Physics of Life Reviews, vol. 2, no. 4, pp. 353–373, 2005.
View at: Publisher Site | Google Scholar
R. Eberhart and J. Kennedy, “A new optimizer using particle swarm theory,” in MHS’95. Proceedings of the Sixth International Symposium on Micro Machine and Human Science, pp. 39–43, Nagoya, Japan, October 1995.
View at: Google Scholar
K. E. Parsopoulos and M. N. Vrahatis, “Recent approaches to global optimization problems through particle swarm optimization,” Natural Computing, vol. 1, no. 2-3, pp. 235–306, 2002.
View at: Publisher Site | Google Scholar
D. Pham, A. Ghanbarzadeh, E. Koc, S. Otri, S. Rahim, and M. Zaidi, The Bees Algorithm, Technical Note, Manufacturing Engineering Centre, Cardiff University, Cardiff, UK, 2005.
J. R. Rohan, C. N. Prasad, and P. Sadiq, “An introduction to the collective behaviour of swarm intelligence,” in Proceedings of the National Conference on Contemporary Research and Innovations in Computer Science, Hyderabad, India, December 2017.
View at: Google Scholar
Y. Shi and R. Eberhart, “A modified particle swarm optimizer,” in 1998 IEEE International Conference on Evolutionary Computation Proceedings, pp. 69–73, IEEE World Congress on Computational Intelligence, Anchorage, AK, USA, May 1998.
View at: Google Scholar
Y. M. Xie and G. P. Steven, “A simple evolutionary procedure for structural optimization,” Computers & Structures, vol. 49, no. 5, pp. 885–896, 1993.
View at: Publisher Site | Google Scholar
C. Grosan and A. Abraham, “Hybrid evolutionary algorithms: methodologies, architectures, and reviews,” 2007.
View at: Google Scholar
W. Khan, Hybrid Multiobjective Evolutionary Algorithm Based on Decomposition, PhD, Department of Mathematical Sciences, University of Essex, Colchester, UK, 2012.
W. K. Mashwani, “Hybrid multiobjective evolutionary algorithms: a survey of the state-of-the-art,” International Journal of Computer Science Issues, vol. 8, no. 6, pp. 374–392, 2011.
View at: Google Scholar
W. K. Mashwani, “MOEA/D with DE and PSO: MOEA/D-DE+PSO,” in The Thirty-First SGAI International Conference on Innovative Techniques and Applications of Artificial Intelligence, pp. 217–221, Cambridge, UK, December, 2011.
View at: Google Scholar
W. K. Mashwani, “Comprehensive survey of the hybrid evolutionary algorithms,” International Journal of Applied Evolutionary Computation (IJAEC), vol. 4, no. 2, pp. 1–19, 2013.
View at: Google Scholar
W. K. Mashwani and A. Salhi, “A decomposition-based hybrid multiobjective evolutionary algorithm with dynamic resource allocation,” Applied Soft Computing, vol. 12, no. 9, pp. 2765–2780, 2012.
View at: Google Scholar
W. K. Mashwani, “Multiobjective memetic algorithm based on decomposition,” Applied Soft Computing, vol. 21, pp. 221–243, 2014.
View at: Google Scholar
X. Qian, X. Wang, Y. Su, and L. He, “An effective hybrid evolutionary algorithm for solving the numerical optimization problems,” Journal of Physics: Conference Series, vol. 1004, 2018.
View at: Publisher Site | Google Scholar
J. H. Holland, “Genetic algorithms and the optimal allocation of trials,” SIAM Journal on Computing, vol. 2, no. 2, pp. 88–105, 1973.
View at: Publisher Site | Google Scholar
F. H. M. T. Back and H. Schwefel, “A survey of evolution strategies,” in Proceeding of the Fourth International Conference on “Genetic Algorithms, Morgan Kauffman, Mateo, CA, USA San Diego, CA, USA, July 1991.
View at: Google Scholar
J. R. Koza, Genetic Programming II, MIT press, Cambridge, UK, 1994.
R. Storn and K. V. Price, “Differential evolution-a simple and efficient heuristic for global optimization over continuous spaces,” Tech. Rep., ICSI, New Delhi, India, 1995, Technical Report TR-95-012.
View at: Google Scholar
J. Kim, T. Sharma, B. Kumar, G. Tomar, K. Berry, and W. Hyung, “Research article intercluster Ant colony optimization algorithm for wireless sensor network in dense environment,” International Journal of Distributed Sensor Networks, vol. 151, 2014.
View at: Google Scholar
L. Perez Caceres, M. Lopez-Ibanez, and T. Stutzle, “Ant colony optimization on a limited budget of evaluations,” Swarm Intellingence, vol. 9, 2015.
View at: Publisher Site | Google Scholar
X.-S. Yang and S. Deb, “Cuckoo search via levy flights,” in Proceedings of the 2009 World Congress on Nature & Biologically Inspired Computing, pp. 210–214, IEEE, Kitakyushu, Japan, December 2009.
View at: Google Scholar
R. V. Rao and V. Patel, “An elitist teaching-learning-based optimization algorithm for solving complex constrained optimization problems,” International Journal of Industrial Engineering Computations, vol. 3, no. 4, pp. 535–560, 2012.
View at: Publisher Site | Google Scholar
R. V. Rao, V. J. Savsani, and D. P. Vakharia, “Teaching-learning-based optimization: a novel method for constrained mechanical design optimization problems,” Computer-Aided Design, vol. 43, no. 3, pp. 303–315, 2011.
View at: Publisher Site | Google Scholar
A. Biswas, S. Dasgupta, B. K. Panigrahi et al., “Economic load dispatch using a chemotactic differential evolution algorithm,” in Proceedings of the 4th International Conference on Hybrid Artificial Intelligence Systems, ser. HAIS ’09, pp. 252–260, León, Spain, September 2009.
View at: Google Scholar
S. Chakraverty, D. M. Sahoo, and N. R. Mahato, Concepts of Soft Computing-Fuzzy and ANN with Programming, Springer, Berlin, Germany, 2019.
P. N. Sugunthan, “Problem definitions and evaluation criteria for the CEC 2005 special session on real-parameter optimization,” Natural Computing, vol. 341, 2005.
View at: Google Scholar
O. Kramer, D. E. Ciaurri, and S. Koziel, Derivative-Free Optimization, Springer Berlin Heidelberg, Berlin, Heidelberg, 2011.
R. Storn and K. Price, “Minimizing the real functions of the ICEC’96 contest by differential evolution,” in Proceedings of IEEE International Conference on Evolutionary Computation, pp. 842–844, IEEE, Nagoya, Japan, May 1996.
View at: Google Scholar
R. Storn and K. Price, “Differential evolution–a simple and efficient heuristic for global optimization over continuous spaces,” Journal of Global Optimization, vol. 11, no. 4, pp. 341–359, 1997.
View at: Publisher Site | Google Scholar
U. K. Chakraborty, Advances in Differential Evolution, Springer Publishing Company, Berlin, Germany, 1st edition, 2008.
S. M. Islam, S. Das, S. Ghosh, S. Roy, and P. N. Suganthan, “An adaptive differential evolution algorithm with novel mutation and crossover strategies for global numerical optimization,” IEEE Transactions on Systems, Man, and Cybernetics, Part.B, vol. 42, no. 2, pp. 482–500, 2012.
View at: Google Scholar
R. A. Khanum, M. A. Jan, W. K. Mashwani, N. M. Tairan, H. U. Khan, and H. Shah, “On the hybridization of global and local search methods,” Journal of Intelligent & Fuzzy Systems, vol. 35, no. 3, pp. 3451–3464, 2018.
View at: Publisher Site | Google Scholar
W. K. Mashwani, “Enhanced versions of differential evolution: state-of-the-art survey,” International Journal Computing Sciences and Mathematics, vol. 5, no. 2, pp. 107–126, 2014.
View at: Google Scholar
F. Neri and V. Tirronen, “Recent advances in differential evolution: a survey and experimental analysis,” Artificial Intelligence Review, vol. 33, pp. 61–106, 2010.
View at: Google Scholar
N. Noman and H. Iba, “Accelerating differential evolution using an adaptive local search,” IEEE Transactions On Evolutionary Computation, vol. 12, no. 1, pp. 107–125, 2008.
View at: Publisher Site | Google Scholar
K. Price, R. M. Storn, and J. A. Lampinen, Differential Evolution: A Practical Approach to Global Optimization (Natural Computing Series), Springer-Verlag, Berlin, Germany, 2005.
M. S. Rainer, Differential Evolution: A Practical Approach to Global Optimization (Natural Computing Series), Springer-Verlag, Berlin, Germany, 2005.
J. Zhang and A. C. Sanderson, “JADE: adaptive differential evolution with optional external archive,” IEEE Transactions on Evolutionary Computation, vol. 13, no. 5, pp. 945–958, 2009.
View at: Google Scholar
F. G. Lobo, C. F. Lima, and Z. Michalewicz, Parameter Setting in Evolutionary Algorithms, Springer Publishing Company, Berlin, Germany, 1st edition, 2007.
J. Liang, B. Qu, P. Suganthan, and A. Hernández-Díaz, Problem Definitions and Evaluation Criteria for the CEC 2013 Special Session on Real-Parameter Optimization, Computational Intelligence Laboratory, Zhengzhou University, Zhengzhou, China, 2013.
W. K. Mashwani, S. N. A. Shah, S. B. Belhaouari, and A. Hamdi, “Ameliorated ensemble strategy-based evolutionary algorithm with dynamic resources allocations,” International Journal of Computational Intelligence Systems, vol. 14, pp. 412–437, 2020.
View at: Google Scholar

Copyright

Copyright © 2021 Wali Khan Mashwani et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

802

Downloads

1092

Citations