An effective hybrid firefly algorithm with harmony search for global numerical optimization.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Metaheuristic Computing

Background:

• Metaheuristic algorithms are crucial for solving complex optimization problems.
• Harmony Search (HS) and Firefly Algorithm (FA) are effective but have limitations.
• Hybridization offers a potential pathway to enhance algorithmic performance.

Purpose of the Study:

• To propose a novel hybrid metaheuristic algorithm, HS/FA, for function optimization.
• To leverage the exploration strengths of HS and the exploitation capabilities of FA.
• To improve convergence speed and solution accuracy compared to individual algorithms.

Main Methods:

• A hybrid approach combining Harmony Search (HS) and Firefly Algorithm (FA) was developed.
• The HS/FA algorithm integrates HS exploration with FA exploitation.
• A 'top fireflies' scheme was introduced to reduce computation time.
• Harmony Search was used for mutation during firefly updates.

Main Results:

• The HS/FA algorithm demonstrated faster convergence than standard HS and FA.
• Experimental results on various benchmark functions validated the algorithm's effectiveness.
• HS/FA outperformed the standard FA and eight other optimization methods.
• The 'top fireflies' scheme effectively reduced the running time.

Conclusions:

• The proposed HS/FA hybrid metaheuristic is a superior optimization method.
• Hybridization effectively combines the strengths of HS and FA for improved performance.
• HS/FA offers a promising approach for complex function optimization tasks.