A Multi-Strategy Adaptive Hybrid Optimization Algorithm for Benchmark Functions and Engineering Design Problems.

Area of Science:

• Engineering optimization
• Computational intelligence
• Metaheuristic algorithms

Background:

• Engineering optimization problems are becoming more complex, demanding advanced methods for global optima.
• Existing algorithms often have limitations in their update mechanisms and decay schedules.

Purpose of the Study:

• Introduce the Adaptive Hybrid Optimization (AHO) algorithm.
• Address limitations of single-equation updates and linear decay schedules in optimization.
• Enhance exploration and exploitation balance for complex engineering problems.

Main Methods:

• AHO employs four distinct, randomly selected position-update strategies for continuous diversity.
• A dual-leader probabilistic guidance mechanism uses asymmetric weighting between top solutions.
• A nonlinear power-changing control parameter D replaces linear decay schedules for better exploration/exploitation balance.

Main Results:

• AHO ranked first among eight algorithms on 23 CEC2005 benchmark functions (best overall rank 1.78).
• Statistically significant improvements were observed over seven competitor algorithms on benchmark and constrained engineering problems.
• AHO achieved top rankings across five metrics for four truss optimization problems.

Conclusions:

• The Adaptive Hybrid Optimization (AHO) algorithm offers a superior approach to complex engineering optimization.
• AHO demonstrates enhanced performance through its novel structural contributions and control parameter.
• The algorithm provides competitive convergence and solution quality compared to established methods.