A Firefly Algorithm for Mixed-Variable Optimization Based on Hybrid Distance Modeling
Summary: arXiv:2603.26792v1 Announce Type: cross
Introduction
In the realm of optimization, many real-world problems present challenges that involve mixed-variable search spaces. These spaces consist of different types of decision variables, including continuous, ordinal, and categorical. Traditional population-based metaheuristic algorithms are typically tailored for either continuous or discrete optimization problems, leading to inefficiencies when applied to heterogeneous variable types. This paper introduces an innovative adaptation of the Firefly Algorithm specifically designed for mixed-variable optimization problems, referred to as FAmv.
Methodology
The core of the proposed method, FAmv, lies in its modified distance-based attractiveness mechanism. This mechanism effectively integrates both continuous and discrete components into a single, unified formulation. The use of a mixed-distance approach allows for a more accurate modeling of heterogeneous search spaces, striking an essential balance between exploration and exploitation during the optimization process.
Performance Evaluation
To assess the efficacy of the FAmv algorithm, extensive evaluations were conducted using the CEC2013 mixed-variable benchmark. This benchmark encompasses a diverse range of functions, including:
- Unimodal functions
- Multimodal functions
- Composition functions
The results from these evaluations indicate that FAmv not only achieves competitive performance but often surpasses various state-of-the-art mixed-variable optimization algorithms. The effectiveness of the proposed algorithm reflects its ability to navigate complex search spaces efficiently, adapting to the intrinsic characteristics of the variable types involved.
Applications
Beyond benchmark testing, the applicability of FAmv extends to practical scenarios, particularly in the field of engineering design. The experiments conducted on various engineering design problems further underscore the robustness of the proposed approach. The results illustrate that FAmv can effectively tackle complex optimization tasks, providing reliable solutions to real-world challenges.
Conclusion
This study highlights the significance of incorporating appropriate distance formulations into the Firefly Algorithm as a viable strategy for solving complex mixed-variable optimization problems. The findings suggest that FAmv not only enhances the algorithm’s adaptability to heterogeneous search spaces but also improves its overall performance across a variety of optimization contexts. As researchers continue to explore the potential of hybrid algorithms, FAmv represents a promising advancement in the field of optimization.
Future Work
Future research may focus on further refining the FAmv algorithm, exploring additional hybridization strategies, and expanding its applicability to an even broader range of optimization problems. Additionally, collaborative efforts in the community could enhance the robustness of mixed-variable optimization techniques, paving the way for innovative solutions in diverse fields.