Regenerative Population Strategy-I (RPS-I)

This repository contains Python implementations of RPS-I for multiple metaheuristic algorithms, illustrating how RPS-I dynamically addresses structural bias during optimization.

The code includes

• 6 files (e.g. RPS_I_GA) for comparison of strauctual bias with the base version through Generalised Signature Test
• 6 files (e.g. RPS_I_GA_POPULATION_PLOT) for visuliazation of popualtion
• Tension/Compression Spring Design Problem (continuous convex benchmark)
• Pressure Vessel Design Problem (engineering design benchmark)

How to Run

• Install Dependencies: Python 3.7+, NumPy, Matplotlib (for plotting convergence curves)
• Clone or Download the repository.
• Navigate to the repository folder, then run either problem’s script (e.g. for the spring design)

RPS-I Overview

Regenerative Population Strategy-I is a dynamic approach for mitigating structural bias. At each generation, measure:

• Population diversity (α)
• Improvement rate (β)
• Compute γ
• Reinitialize N indvidulas based on Eq. (14) This helps the algorithm avoid premature convergence and maintain better exploration/exploitation trade-offs.

Citation

If you use or reference this code in your publications, please cite the paper:

Kanchan Rajwar et al., “Regenerative Population Strategy-I: A Dynamic Methodology to Mitigate Structural Bias in Metaheuristic Algorithms.”

License

This code is provided for academic and research purposes.

Contact

For questions or collaboration, feel free to contact:

• Author: Kanchan Rajwar
• Email: kanchanrajwar1519@gmail.com