Multi-strategy enhanced sand cat swarm optimization algorithm and its engineering applications.

Area of Science:

• Optimization Algorithms
• Swarm Intelligence
• Computational Intelligence

Background:

• The standard sand cat swarm optimization (SCSO) algorithm faces challenges with slow convergence and getting trapped in local optima.
• Enhancing swarm intelligence algorithms is crucial for improving their efficiency and applicability in complex problem-solving.

Purpose of the Study:

• To propose a multi-strategy enhanced sand cat swarm optimization (MESCSO) algorithm.
• To improve the convergence speed and local optima escape ability of the SCSO algorithm.

Main Methods:

• Population initialization using improved sine mapping with random opposition-based learning (ISMROBL).
• Dynamic balancing of global exploration and local exploitation via a nonlinear decreasing parameter.
• Incorporation of generalized quadratic interpolation (GQI) for global search and improved mean differential mutation (IMDM) for local exploitation.
• Application of accelerated opposition-based learning (AOBL) to refine solutions and enhance local optima escape.

Main Results:

• MESCSO demonstrated superior performance over nine other algorithms on 23 standard and CEC2014 benchmark functions.
• Applied to five constrained engineering design problems, MESCSO showed significant improvements over the original SCSO algorithm.
• Specific performance gains on engineering problems ranged from 0.03% to 1.47%.

Conclusions:

• The proposed MESCSO algorithm effectively enhances convergence efficiency and the ability to escape local optima.
• MESCSO exhibits strong performance and applicability for solving complex optimization problems, including constrained engineering designs.