An Enhanced Snow Geese Optimizer Integrating Multiple Strategies for Numerical Optimization.

Area of Science:

• Artificial Intelligence
• Optimization Algorithms
• Computational Intelligence

Background:

• The standard snow geese algorithm (SGA) suffers from limited population diversity and unbalanced search tendencies.
• These limitations hinder its effectiveness in complex optimization tasks.

Purpose of the Study:

• To introduce an enhanced snow geese algorithm (ESGA) that addresses the shortcomings of the standard SGA.
• To improve population diversity, balance exploration and exploitation, and enhance the ability to escape local optima.

Main Methods:

• An adaptive switching strategy was employed to dynamically balance exploitation and exploration.
• A dominant group guidance strategy was introduced to improve overall population quality.
• A dominant stochastic difference search strategy was designed to enrich diversity and facilitate escape from local optima.
• Ablation experiments on the CEC2017 test set validated the effectiveness of individual improvements.
• Comparative studies on the CEC2022 test suite benchmarked ESGA against state-of-the-art algorithms.

Main Results:

• Ablation studies confirmed the positive contribution of each proposed strategy to the algorithm's performance.
• The ESGA demonstrated superior performance compared to highly cited and powerful improved algorithms on the CEC2022 test suite.
• The algorithm successfully solved the complex robot path planning problem, showcasing its practical applicability.

Conclusions:

• The proposed ESGA effectively overcomes the limitations of the standard SGA by enhancing population diversity and search capabilities.
• ESGA offers a robust and efficient optimization approach, outperforming existing methods on benchmark test functions.
• The algorithm shows significant potential for solving real-world complex problems, including path planning.