Aquila Optimiser Optimised

Overview

Implementations and analyses of optimization algorithms centred on the Aquila Optimizer (AO), a metaheuristic inspired by the hunting strategies of Aquilas, together with a modified variant developed in this work. The repository compares these algorithms on benchmark functions with statistical evaluation, and applies the modified variant to classical engineering design problems.

What it does

• Algorithm implementations (7algos/): the Aquila Optimizer (AO), Equilibrium Optimizer (EO), Grasshopper Optimization Algorithm (GOA) and Particle Swarm Optimization (PSO), along with comparison graphs across all benchmark functions.
• Modified Aquila variant (MVAO/): a Jupyter notebook demonstrating the modified optimizer, plus the accompanying research paper (AQUILA_OPTIMISER_2025).
• Variant comparison (Variants_Comparison/): a script comparing the modified variants, with Wilcoxon test results stored as CSV.
• Benchmark evaluation (matrixEval_vs_cec/): notes on the benchmark evaluations and Wilcoxon results for all function pairs.
• Engineering problems (mvao_on_engineering_problems/): the modified optimizer applied to clutch brake design, speed reducer design, pressure vessel design, spring design and truss design.
• Reference material (original_aquila/): the original Aquila optimizer code and the base paper.

How it's built

Pure Python, depending only on numpy, matplotlib and scipy. Each algorithm is a standalone script (for example python AO.py), and the statistical comparisons are produced as CSV files. Licensed under MIT.

Results

The README does not quote numerical results; the Wilcoxon test outputs are provided as CSV files in the repository rather than summarised in prose.