A compact compound sinusoidal differential evolution algorithm for solving optimisation problems in memory-constrained environments. (30th December 2021)

• Record Type: Journal Article
• Authors: Khalfi, Souheila; Draa, Amer; Iacca, Giovanni
• Is Part Of: Expert systems with applications. Volume 186(2021)
• Publication Date: 2021-12-30
• Subjects: Limited-memory hardware -- Compact optimisation -- Evolutionary algorithms -- Parameter setting -- Compound Sinusoidal Differential Evolution; Expert systems (Computer science) -- Periodicals; Systèmes experts (Informatique) -- Périodiques; Electronic journals; 006.33
• DOI: 10.1016/j.eswa.2021.115705
• Languages: English
• ISSNs: 0957-4174

Abstract

Abstract: In this paper, a new compact algorithm is proposed. Two sinusoidal formulas are used to automatically adjust the crossover rate and the mutation scaling factor in the compact Differential Evolution (cDE) metaheuristic. The proposed algorithm, called Compound Sinusoidal cDE, CScDE, is compared to seven state-of-the-art compact algorithms on the well-known BBOB test-bed, the CEC-2014 test suite for continuous optimisation, as well as five real-world optimisation problems chosen from the CEC-2011 benchmarks. The CScDE algorithm outperformed its competitors for most problem categories and over most dimensions. It is also compared to some well-established population-based metaheuristics. Highlights: A compact compound sinusoidal differential evolution, CScDE, is proposed. CScDE uses sinusoidal formulas to adjust the parameters of compact DE. Both binomial and exponential crossover are tested. CScDE is applied on 89 reference problems with several dimensions. CScDE outperformed state-of-the-art compact algorithms.