A Modified Complex-Valued Encoding Greater Cane Rat Algorithm for Global Optimization and Constrained Engineering...

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Bio-inspired Computing

Background:

• The standard Greater Cane Rat Algorithm (GCRA) faces challenges like diversity collapse and slow convergence.
• Existing optimization methods struggle with high-dimensional problems and local optima.

Purpose of the Study:

• To introduce a modified complex-valued encoding GCRA (CGCRA) to overcome the limitations of the original GCRA.
• To leverage complex number properties for improved exploration and exploitation in optimization tasks.

Main Methods:

• Developed a complex-valued encoding strategy mapping decision variables onto the complex plane.
• Utilized the real part for local exploitation and the imaginary part for global exploration via phase rotation.
• Implemented a dual-encoding redundancy mechanism for enhanced error tolerance and population heterogeneity.

Main Results:

• The CGCRA demonstrated superior performance across 23 benchmark functions and 12 engineering designs.
• Significant improvements were observed in convergence speed, solution precision, and robustness against stagnation.
• The enhanced algorithm effectively balanced exploration and exploitation, reducing result dispersion.

Conclusions:

• The CGCRA offers a robust and efficient optimization framework by integrating complex number mathematics.
• The proposed method shows significant potential for solving complex, high-dimensional optimization problems in engineering and beyond.
• CGCRA provides enhanced stability, adaptability, and computational efficiency compared to traditional methods.