Abstract
Evolution strategies are a class of general optimisation algorithms which arc applicable to functions that arc multi-modal, nondifferentiable, or even discontinuous. Although recombination operators have been introduced into evolution strategies, the primary search operator is still mutation. Classical evolution strategies rely on Gaussian mutations. A new mutation operator based on the Cauchy distribution is proposed in this paper. It is shown empirically that the new evolution strategy based on Cauchy mutation outperforms the classical evolution strategy on most of the 23 benchmark problems tested in this paper. The paper also shows empirically that changing the order of mutating the objective variables and mutating the strategy parameters docs not alter the previous conclusion significantly, and that Cauchy mutations with different scaling parameters still outperform the Gaussian mutation with self-adaptation. However, the advantage of Cauchy mutations disappears when recombination is used in evolution strategies. It is argued that the search step size plays an important role in determining evolution strategies' performance. The large step size of recombination plays a similar role as Cauchy mutation.
Original language | English |
---|---|
Pages (from-to) | 466-491 |
Number of pages | 26 |
Journal | Control and Cybernetics |
Volume | 26 |
Issue number | 3 |
Publication status | Published - 1997 |
Externally published | Yes |
Keywords
- Cauchy mutation
- Evolutionary strategies
- Function optimisation