Test package for Metaheuristic optimisers

A comprehensive benchmarking framework for evaluating and comparing metaheuristic optimization algorithms across mathematical test functions

This script serves as a complete evaluation and benchmarking suite for comparing the performance of multiple metaheuristic optimization algorithms (SA, HS, PSO, BA, GWO, DE, and IUO).

test_cec2017s.m is the main file.

The framework assesses algorithms across two distinct problem categories:

• Mathematical Benchmark Functions: Tests performance on 23 standard benchmark functions (F1 to F23).
• Engineering Design Problems: Solves 6 real-world constrained design cases, including Cantilever Beam, Pressure Vessel Design, Gear Train Design, Three Bar Truss, String Design, and Welded Beam.
• Performs multiple independent runs (default: 30) to compute robust statistics, including Mean, Standard Deviation, Minimum, Maximum, and Average Computation Time.
• Visualizations: Automatically generates and highlights semi-log convergence curves for each function and problem.
• Data Export: Automatically aggregates and saves all statistical results into structured CSV files (BenchmarkStatistics.csv and EngineeringStatistics.csv) for easy reporting and analysis.

IUO: Intelligence Unit-Inspired Hybrid Metaheuristic Optimiser[v1] | Preprints.org