Abstract
The RIME (Rime-ice) optimization algorithm is a physics-based metaheuristic inspired by the crystallization dynamics of rime ice. Since its introduction in 2023, nearly 170 variants and applications have appeared across more than 40 journals, yet no dedicated survey exists to organize this rapidly expanding literature. This paper fills that gap by presenting the first comprehensive survey of the RIME algorithm family, covering 169 studies published between January 2023 and December 2025. We (1) formalize the theoretical foundations of RIME, establishing a clear mapping from soft-rime branching to global exploration and from hard-rime densification to local exploitation; (2) construct a six-dimensional taxonomy of enhancement strategies encompassing population initialization, exploration enhancement, exploitation refinement, exploration–exploitation balancing, constraint handling, and multi-objective/distributed extensions; (3) systematically review engineering applications across seven domains including robot path planning, unmanned systems, engineering optimization, feature selection, image processing, wireless networks, and emerging applications; (4) synthesize benchmark results on CEC test suites through cross-study performance aggregation with statistical significance analysis; (5) position RIME within the broader metaheuristic landscape through comparative analysis with 19 peer algorithms; and (6) identify open challenges and future research directions spanning convergence theory, high-dimensional scaling, distributed optimization, and integration with deep reinforcement learning and large language models. A structured research roadmap is provided to guide both algorithm designers and practitioners. No prior dedicated survey of the RIME algorithm family exists in the literature; we hope this work serves as a useful reference for researchers and engineers entering this rapidly evolving field.

This publication has 40 references indexed in Scilit: