Orthogonal pinhole-imaging-based learning salp swarm algorithm with self-adaptive structure for global optimization.

Area of Science:

• Optimization Algorithms
• Computational Intelligence
• Bio-inspired Computing

Background:

• The Salp Swarm Algorithm (SSA) is a popular bio-inspired metaheuristic for global optimization.
• Existing SSA variants often struggle with local optima and balancing exploration/exploitation.
• Enhancing SSA's efficiency and convergence is crucial for complex optimization problems.

Purpose of the Study:

• To propose a novel enhanced Salp Swarm Algorithm (OPLSSA) for improved global optimization.
• To introduce a pinhole-imaging-based learning (PIBL) strategy combined with orthogonal experimental design (OED) for escaping local optima.
• To develop an adaptive parameter control method for better exploration-exploitation balance.

Main Methods:

• Developing a Pinhole-Imaging-based Learning (PIBL) strategy.
• Integrating PIBL with Orthogonal Experimental Design (OED) to create the OPIBL mechanism.
• Designing a novel adaptive conversion parameter for exploration-exploitation balance.
• Conducting comparative experiments on 23 benchmark functions and 30 IEEE CEC2017 problems.

Main Results:

• The proposed OPLSSA algorithm demonstrates significant improvements over existing well-established algorithms.
• OPLSSA shows enhanced ability to escape local optima.
• The adaptive parameter method effectively balances exploration and exploitation capabilities.
• OPLSSA achieved superior performance across a majority of tested benchmark functions.

Conclusions:

• The OPLSSA algorithm offers a promising enhancement to the Salp Swarm Algorithm.
• The novel PIBL and OED integration effectively addresses local optima stagnation.
• The adaptive parameter control contributes to robust optimization performance.
• OPLSSA represents a valuable contribution to the field of global optimization algorithms.