Differential evolution with individual-dependent and dynamic parameter adjustment
- 255 Downloads
Differential evolution (DE) is a powerful and versatile evolutionary algorithm for global optimization over continuous search space, whose performance is significantly influenced by its mutation operator and control parameters (population size, scaling factor and crossover rate). In order to enhance the performance of DE, we adopt a new variant of classic mutation operator, a gradual decrease rule for population size, an individual-dependent and dynamic strategy to generate the required values of scaling factor and crossover rate during the evolutionary process, respectively. In the proposed variant of DE (denoted by IDDE), the adopted mutation operator merges the superiority of two classic mutation operators (DE/best/2 and DE/rand/2) together, and the adjustment mechanism of control parameters applies the fitness value information of each individual and dynamic fluctuation rule, which can provide a better balance between the exploration ability and exploitation ability. To verify the performance of proposed IDDE, a suite of thirty benchmark functions is applied to conduct the simulation experiment. The simulation results demonstrate that the proposed IDDE performs significantly better than five state-of-the-art DE variants and other two evolutionary algorithms.
KeywordsDifferential evolution Individual-dependent strategy Dynamic parameter adjustment Evolutionary algorithms Global optimization
Compliance with ethical standards
Conflict of interest
All authors have no conflict of interest.
Human participants or animals
This article does not contain any studies with human participants or animals performed by any of the authors.
- Das S, Suganthan PN (2011b) Problem definitions and evaluation criteria for CEC 2011 competition on testing evolutionary algorithms on real world optimization problems. Jadavpur University, Kolkata, India, and Nanyang Technological University, Singapore 2010Google Scholar
- Kennedy J, Eberhart R, Shi Y (2001) Swarm intelligence. Morgan Kaufman, San FranciscoGoogle Scholar
- Liang JJ, Qu BY, Suganthan PN (2013) Problem definitions and evaluation criteria for the CEC 2014 special session and competition on single objective real-parameter numerical optimization. Zhengzhou University, China, and Nanyang Technological University, SingaporeGoogle Scholar