A-DVM: A Self-Adaptive Variable Matrix Decision Variable Selection Scheme for Multimodal Problems

Area of Science:

• Computational Intelligence
• Swarm Intelligence
• Optimization Algorithms

Background:

• The Artificial Bee Colony (ABC) algorithm is a versatile swarm intelligence technique.
• Stochastic decision variable selection in ABC can limit local search capabilities.
• Balancing exploration and exploitation is crucial for effective optimization.

Purpose of the Study:

• To propose a self-adaptive mechanism for decision variable selection in the ABC algorithm.
• To enhance the local search capability and overall performance of the ABC algorithm.
• To introduce the Adaptive Decision Variable Matrix (A-DVM) for improved exploration-exploitation balance.

Main Methods:

• Developed the Adaptive Decision Variable Matrix (A-DVM) incorporating both stochastic and deterministic selection.
• A-DVM dynamically regulates selection based on solution sparsity estimation.
• Validated the approach on 15 highly multimodal benchmark optimization problems.

Main Results:

• The proposed A-DVM mechanism demonstrated improved performance and robustness for the ABC algorithm.
• Comparative analysis showed superior results against standard ABC, its variants, and other population-based algorithms.
• Significant performance gains were observed on challenging optimization instances.

Conclusions:

• The A-DVM is an effective enhancement for the ABC algorithm, improving its optimization capabilities.
• The adaptive mechanism successfully balances exploration and exploitation for better results.
• This approach offers a promising direction for advancing swarm intelligence optimization techniques.