Adaptive dynamic crayfish algorithm with multi-enhanced strategy for global high-dimensional optimization and...

Area of Science:

• Computational Intelligence
• Metaheuristic Optimization
• Algorithm Development

Background:

• The Crayfish Optimization Algorithm (COA) faces challenges with slow convergence and local optima.
• Existing metaheuristic algorithms often struggle with balancing exploration and exploitation.

Purpose of the Study:

• To introduce an improved COA variant, Adaptive Dynamic COA with a Locally enhanced escape operator (AD-COA-L).
• To address the limitations of poor convergence speed and local optimum convergence in the original COA.

Main Methods:

• Utilizing Bernoulli map initialization for a high-quality, evenly distributed initial population.
• Applying Adaptive Lens Opposition-Based Learning (ALOBL) to escape local optima and enhance solution quality.
• Incorporating a local escape operator (LEO) to promote information sharing and prevent isolated solutions.
• Introducing a novel inertia weight to balance exploration and exploitation capabilities.

Main Results:

• AD-COA-L demonstrated superior accuracy and balanced exploration-exploitation compared to 18 other algorithms on 29 CEC2017 benchmark functions.
• The algorithm showed improved convergence speed across various dimensions (50 and 100).
• AD-COA-L proved effective in solving seven complex real-world engineering optimization problems.

Conclusions:

• AD-COA-L significantly outperforms existing algorithms in terms of accuracy, convergence, and solution quality.
• The proposed enhancements effectively mitigate local optima convergence and improve overall optimization performance.
• AD-COA-L presents a competitive and advantageous metaheuristic for diverse optimization challenges.