Many-objective BAT algorithm

Area of Science:

• Optimization Algorithms
• Computational Intelligence
• Evolutionary Computation

Background:

• Many-objective optimization problems (MaOPs) present challenges in achieving both diversity and convergence in Pareto approximations.
• Existing multi-objective evolutionary algorithms (MOEAs) and reference point techniques address these challenges with varying success.

Purpose of the Study:

• To develop an effective algorithm for handling many-objective optimization problems.
• To enhance the diversity and convergence of Pareto fronts in MaOPs.

Main Methods:

• Adaptation of the bat algorithm, termed Many Objective Bat Algorithm (MaOBAT), inspired by bat echolocation.
• Utilization of dominance rank for solution selection and dynamically allocated reference points to guide the search process.

Main Results:

• The proposed MaOBAT algorithm demonstrates significant advantages over state-of-the-art algorithms.
• Experimental results indicate improved quality of solutions in terms of convergence and diversity.

Conclusions:

• MaOBAT effectively addresses the challenges of diversity and convergence in many-objective optimization problems.
• The algorithm's biologically inspired approach and use of reference points contribute to superior performance.