RDEx-CSOP: Feasibility-Aware Reconstructed Differential Evolution with Adaptive epsilon-Constraint Ranking
Summary: arXiv:2603.27090v1 Announce Type: cross
Abstract
Constrained single-objective numerical optimisation requires both feasibility maintenance and strong objective-value convergence under limited evaluation budgets. This report documents RDEx-CSOP, a constrained differential evolution variant used in the IEEE CEC 2025 numerical optimisation competition (C06 special session). RDEx-CSOP combines success-history parameter adaptation with an exploitation-biased hybrid search and an ε-constraint handling mechanism with a time-varying threshold. We evaluate RDEx-CSOP on the official CEC 2025 CSOP benchmark using the U-score framework (Speed, Accuracy, and Constraint categories). The results show that RDEx-CSOP achieves the highest total score and the best average rank among all released comparison algorithms, mainly through strong speed and competitive constraint-handling performance across the 28 benchmark functions.
Introduction
The need for efficient algorithms in constrained single-objective numerical optimisation is paramount, especially as the complexity and dimensionality of problems increase. RDEx-CSOP addresses these challenges by integrating advanced techniques that enhance both feasibility and convergence.
Key Features of RDEx-CSOP
- Success-History Parameter Adaptation: This feature allows the algorithm to adapt its parameters based on past successes, thus improving its efficiency in finding optimal solutions.
- Exploitation-Biased Hybrid Search: By focusing on exploiting known good regions of the search space, RDEx-CSOP enhances its convergence rates significantly.
- ε-Constraint Handling Mechanism: The adaptive ε-constraint approach is crucial for managing trade-offs between multiple objectives effectively, particularly under varying conditions.
- Time-Varying Threshold: This mechanism dynamically adjusts the constraints over time, allowing the algorithm to explore new solutions while maintaining feasibility.
Performance Evaluation
The evaluation of RDEx-CSOP was conducted on the CEC 2025 CSOP benchmark, which comprises a series of challenging test functions designed to assess the performance of optimisation algorithms. The U-score framework was employed to measure three main categories:
- Speed: The algorithm’s ability to find solutions quickly without compromising quality.
- Accuracy: The precision of the solutions produced by the algorithm.
- Constraint Handling: The effectiveness of the algorithm in managing constraints while searching for optimal solutions.
Results and Findings
RDEx-CSOP demonstrated exceptional performance, achieving the highest total score and the best average rank among all comparison algorithms. The results highlighted its strengths in:
- Rapid convergence to optimal solutions.
- Robust handling of constraints across the 28 benchmark functions.
- Overall speed, which was found to outperform many existing algorithms.
Conclusion
In summary, RDEx-CSOP represents a significant advancement in constrained single-objective numerical optimisation. Its innovative approach to parameter adaptation, hybrid search techniques, and constraint handling positions it as a leading algorithm in the field. As optimisation challenges continue to evolve, RDEx-CSOP provides a flexible and efficient tool for researchers and practitioners alike.