An adaptive differential evolution algorithm with elite gaussian mutation and bare-bones strategy.

Area of Science:

• Optimization Algorithms
• Computational Intelligence
• Evolutionary Computation

Background:

• Differential evolution (DE) and bare-bones (BB) algorithms offer efficiency but struggle with complex multimodal problems.
• DE excels in global search, while BB is strong in local search, presenting an opportunity for synergistic combination.
• Existing methods require further improvement for effectively tackling challenging optimization tasks.

Purpose of the Study:

• To propose an adaptive differential evolution algorithm (EGBDE) integrating elite Gaussian mutation and bare-bones operations.
• To enhance global search ability and search accuracy in differential evolution.
• To achieve a parameter-free, adaptive approach that balances exploration and exploitation.

Main Methods:

• Developed an adaptive differential evolution algorithm (EGBDE).
• Incorporated an elite Gaussian mutation strategy using information from selected elite individuals.
• Adjusted mean and variance of bare-bones operations adaptively.
• Utilized an adaptive adjustment factor to balance differential mutation and elite Gaussian mutation.

Main Results:

• EGBDE demonstrated enhanced global search ability and improved search accuracy.
• The algorithm effectively balanced exploration and exploitation for complex multimodal problems.
• Comparative tests on twenty functions showed EGBDE outperformed competing algorithms.

Conclusions:

• The proposed EGBDE algorithm effectively combines the strengths of DE and BB algorithms.
• EGBDE offers a robust and adaptive solution for complex multimodal optimization.
• The parameter-free nature and adaptive balancing contribute to its excellent performance.