CSCAHHO: Chaotic hybridization algorithm of the Sine Cosine with Harris Hawk optimization algorithms for solving...

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Metaheuristics

Background:

• The No Free Lunch (NFL) theorem necessitates continuous development of novel optimization algorithms.
• Existing algorithms like Sine Cosine Algorithm (SCA) and Harris Hawk Optimization (HHO) have distinct strengths in convergence and speed.
• Hybridization offers a promising avenue to leverage and enhance these individual algorithm capabilities.

Purpose of the Study:

• To propose a novel hybrid optimization algorithm, CSCAHHO, by integrating SCA and HHO.
• To enhance the exploration-exploitation balance and randomness in optimization through an energy parameter and chaos theory.
• To validate the efficacy of the proposed CSCAHHO algorithm against existing state-of-the-art methods.

Main Methods:

• Hybridization of Sine Cosine Algorithm (SCA) and Harris Hawk Optimization (HHO) into the CSCAHHO algorithm.
• Introduction of an energy parameter to balance exploration and exploitation.
• Incorporation of chaos theory to improve population randomness.
• Redefinition of updating equations by combining SCA and HHO principles.

Main Results:

• The CSCAHHO algorithm demonstrated superior performance across 27 benchmark functions, CEC 2014 functions, and 3 engineering problems.
• Comparative analysis showed CSCAHHO outperformed original SCA, HHO, and other algorithms like AOA, SOA, STOA, AO, and ChOA.
• Key performance indicators included faster convergence rates, reduced residual errors, and improved stability.

Conclusions:

• The proposed CSCAHHO algorithm effectively overcomes limitations of individual algorithms, offering enhanced optimization capabilities.
• The hybridization strategy, incorporating an energy parameter and chaos, significantly improves convergence speed and solution accuracy.
• CSCAHHO presents a robust and efficient metaheuristic for complex optimization tasks, with Matlab code available for reproducibility.