A Novel Sarpa Salpa-Inspired Optimization Algorithm: Performance Evaluation and Comparison with Particle Swarm Optimization on Benchmark Functions
الملخص
This paper introduces a novel optimization algorithm inspired by the behavior of the Sarpa Salpa fish, referred to as SSOA. The algorithm mimics the natural exploration and exploitation strategies of Sarpa Salpa, incorporating adaptive mechanisms for improved search efficiency in complex multimodal landscapes. Performance evaluation is conducted on standard benchmark functions Rastrigin, Griewank, Sphere, and Ackley across multiple dimensionalities (2D, 5D, 10D, 20D, and 50D). Statistical analyses over repeated trials show that SSOA outperforms the classical Particle Swarm Optimization (PSO) algorithm in terms of accuracy, robustness, and success rate, especially in higher-dimensional problems. In addition, a sensitivity analysis of key parameters (alpha, beta, gamma, and inertia weight) demonstrates the algorithm's resilience to parameter variations, while highlighting that extreme parameter values can degrade performance. Despite a moderate increase in computational cost, the algorithm demonstrates strong potential for solving challenging global optimization problems.
كيفية الاقتباس
