Self-adaptive differential evolution algorithm with crossover strategies adaptation and its application in parameter estimation

• Published in: Chemometrics and intelligent laboratory systems Vol. 151; pp. 164 - 171
• Main Authors: Fan, Qinqin, Zhang, Yilian
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 15.02.2016
• Subjects: Crossover strategy adaption; Differential evolution; Evolutionary computation; Parameter estimation; Self-adaptive; Differential evolution; Self-adaptive; Parameter estimation; Evolutionary computation; Crossover strategy adaption
• ISSN: 0169-7439; 1873-3239
• DOI: 10.1016/j.chemolab.2015.12.020

The performance of differential evolution (DE) is significantly influenced by the choice of crossover strategies; therefore, a self-adaptive differential evolution algorithm with crossover strategies adaptation (CSA-SADE) is proposed in this paper to enhance the performance of DE. In CSA-SADE, the suitable control parameters, mutation strategies, and crossover strategies can be achieved in different evolution stages. To demonstrate the effectiveness of CSA-SADE, the proposed algorithm is compared with eight state-of-the-art evolutionary algorithms. The simulation results indicate that CSA-SADE outperforms five improved DE algorithms and three non-DE approaches on a set of 25 CEC2005 benchmark functions. Additionally, the proposed algorithm is employed to estimate the kinetic parameters of mercury oxidation; the results show that CSA-SADE performs better than the compared algorithms in this simulation example. •The crossover strategy can be automatically adjusted in CSA-SADE.•CSA-SADE is compared with five famous DE variants and three non-DEs on 25 CEC2005 test functions.•In CSA-SADE, the appropriate control parameters and mutation strategies can be achieved at different evolution stages.•CSA-SADE is employed to estimate the kinetic parameters of Hg oxidation.