An Efficient Multi-Objective White Shark Algorithm.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Engineering Applications

Background:

• Balancing diversity and stringency in Pareto solutions is crucial for multi-objective optimization.
• Existing algorithms often struggle to achieve both optimal convergence and uniform solution distribution.
• The White Shark Optimization algorithm provides a unique bio-inspired framework for search and exploration.

Purpose of the Study:

• To introduce a novel multi-objective White Shark Optimization algorithm (MONSWSO).
• To enhance the selection of optimal solutions by integrating non-dominated sorting and crowding distance.
• To improve initial population uniformity and adaptive position updating for better exploration.

Main Methods:

• Implemented a chaotic reverse initialization learning strategy for enhanced population uniformity.
• Integrated an elite-guided forgetting mechanism with escape energy and eddy aggregation behavior for adaptive updating.
• Employed non-dominated sorting and crowding distance for optimal solution selection within the population.
• Benchmarked MONSWSO against five state-of-the-art algorithms on 27 multi-objective problems using four key metrics.

Main Results:

• MONSWSO demonstrated superior performance compared to existing multi-objective optimization algorithms.
• The algorithm achieved impressive and satisfactory outcomes across various benchmark functions and project examples.
• Evaluations using inverse generation distance, spatial homogeneity, spatial distribution, and hypervolume confirmed MONSWSO's effectiveness.

Conclusions:

• MONSWSO effectively balances the diversity and stringency of Pareto solutions in multi-objective optimization.
• The proposed algorithm offers a robust and efficient approach for tackling complex optimization challenges.
• MONSWSO shows practical applicability, as evidenced by its successful application in optimizing subway tunnel foundation pit design.