Hierarchical guided manta ray foraging optimization for global continuous optimization problems and parameter...

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Metaheuristic Computing

Background:

• The Manta Ray Foraging Optimization (MRFO) algorithm is effective for engineering problems but suffers from local optima and poor exploration-exploitation balance.
• Fixed parameters and a somersault mechanism focusing on the current best solution limit MRFO's adaptability.

Purpose of the Study:

• To propose an improved optimization algorithm, Hierarchical Guided Manta Ray Foraging Optimization (HGMRFO).
• To address the limitations of MRFO, specifically its tendency to get trapped in local optima and its inadequate balance between exploration and exploitation.

Main Methods:

• Introduced an adaptive somersault factor to dynamically balance exploration and exploitation.
• Developed a novel hierarchical guidance mechanism within the somersault foraging strategy to direct population search.
• Validated HGMRFO against seven state-of-the-art algorithms on IEEE CEC2017 benchmark functions and IEEE CEC2011 real-world problems.

Main Results:

• HGMRFO achieved an average win rate of 73.15% on 29 IEEE CEC2017 benchmark functions.
• Obtained the most optimal solutions on 22 IEEE CEC2011 real-world optimization problems.
• Demonstrated a 97.62% success rate in parameter estimation for multimodal photovoltaic models.

Conclusions:

• HGMRFO effectively overcomes the local optima problem and improves the exploration-exploitation balance compared to standard MRFO.
• The proposed hierarchical guidance mechanism and adaptive parameters significantly enhance optimization performance.
• HGMRFO shows superior applicability and effectiveness, particularly in solving complex problems like photovoltaic parameter estimation.