Differential evolution algorithm

This library aims to simplify the usage of differential evolution algorithm and enables effective comparison of various versions. Various standard testing and population diversity measurement functions are also implemented. Only binomial crossover is implemented so far, exponential crossover will be implemented soon.

Implemented mutation strategies

• DE/rand/1
• DE/rand/2
• DE/best/1
• DE/best/2
• DE/current-to-best/1
• DE/current-to-best/2
• DE/current-to-rand/1
• DE/current-to-rand/2

Implemented crossover

• binomial

Implemented testing functions

• sphere function
• Rastrigin function
• Beale function
• Booth function
• Matyas function
• Himmelblau's function
• Bukin function
• McCormick function
• Three-hump camel function
• Ackley function
• Goldstein-Price function
• Lévi function n.13
• Easom function
• Eggholder function
• Schaffer function n.2

Implemented functions for population diversity estimation

• standard deviation of fitness function

Implemented initial generation

• random
• oppositional based learning
• tent map
• quasi-oppositional based learning
• Sobol's pseudo-random sequence

Will be implemented soon

• randomized F
• adaptively updated F
• Neighborhood Search Differential Evolution
• trigonometric mutation
• directed mutation
• population size reduction
• scale factor golden section search
• some self-adapting parameters schemes
• exponential crossover
• plotting testing functions