A Modified Sine-Cosine Algorithm Based on Neighborhood Search and Greedy Levy Mutation

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Heuristic Computing

Background:

• The basic sine-cosine algorithm (SCA) suffers from low solution precision and slow convergence, limiting its effectiveness in global optimization.
• Addressing these deficiencies is crucial for advancing computational intelligence and solving complex problems.

Purpose of the Study:

• To propose a novel improved sine-cosine algorithm (SCA) that overcomes the limitations of the basic version.
• To enhance the global exploration and local development capabilities of the SCA for superior optimization performance.

Main Methods:

• Implemented an improved sine-cosine algorithm (SCA) incorporating three key strategies.
• Utilized exponential decreasing conversion parameter and linear decreasing inertia weight for balanced exploration and development.
• Incorporated random individuals near optimal solutions to escape local optima and expand search range.
• Applied a greedy Levy mutation strategy to optimal individuals to improve local development.

Main Results:

• The proposed improved sine-cosine algorithm (SCA) effectively avoids premature convergence into local optima.
• Demonstrated significantly faster convergence speeds compared to the basic SCA.
• Achieved higher optimization accuracy in solving global optimization problems.

Conclusions:

• The enhanced sine-cosine algorithm (SCA) offers a robust solution for global optimization problems.
• The integration of adaptive parameters, random individual replacement, and Levy mutation enhances algorithmic performance.
• This improved SCA provides a valuable tool for researchers and practitioners seeking efficient and accurate optimization solutions.