| _version_ | 1866901930974904320 |
|---|---|
| author | Zhang, Jincheng |
| author_facet | Zhang, Jincheng |
| contents | <p><span>This paper proposes a novel optimization algorithm called Cryo-Entropy Gradient Dynamics (CEGD), which is inspired by three key phenomena in cryogenics and low-temperature physics: phase condensation, dilution flow, and quantum fluctuation and tunneling. The algorithm maps cryogenic physics phenomena to the optimization process, introducing three mechanisms: Dynamic Condensation Potential (DCP), Entropy-Guided Dilution Flow (EGF), and Coherent Quantum Tunneling (CQT), to achieve efficient search and local minima escape capabilities for non-convex optimization problems. The algorithm employs a dual-temperature mechanism: macroscopic temperature controls the global exploration and convergence rate, while microscopic temperature introduces periodic perturbations to simulate quantum fluctuations, thereby enhancing local exploration capabilities while maintaining convergence. This paper provides a detailed description of the algorithm's mathematical structure and presents complete iterative formulas and pseudocode. This method theoretically possesses excellent global search capabilities and adaptive characteristics, providing a new approach for the research of optimization algorithms inspired by low-temperature physics.</span></p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17852914 |
| institution | Zenodo |
| language | |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Cryogenics-Inspired Entropy-Guided Gradient Dynamics for Optimization Zhang, Jincheng <p><span>This paper proposes a novel optimization algorithm called Cryo-Entropy Gradient Dynamics (CEGD), which is inspired by three key phenomena in cryogenics and low-temperature physics: phase condensation, dilution flow, and quantum fluctuation and tunneling. The algorithm maps cryogenic physics phenomena to the optimization process, introducing three mechanisms: Dynamic Condensation Potential (DCP), Entropy-Guided Dilution Flow (EGF), and Coherent Quantum Tunneling (CQT), to achieve efficient search and local minima escape capabilities for non-convex optimization problems. The algorithm employs a dual-temperature mechanism: macroscopic temperature controls the global exploration and convergence rate, while microscopic temperature introduces periodic perturbations to simulate quantum fluctuations, thereby enhancing local exploration capabilities while maintaining convergence. This paper provides a detailed description of the algorithm's mathematical structure and presents complete iterative formulas and pseudocode. This method theoretically possesses excellent global search capabilities and adaptive characteristics, providing a new approach for the research of optimization algorithms inspired by low-temperature physics.</span></p> |
| title | Cryogenics-Inspired Entropy-Guided Gradient Dynamics for Optimization |
| url | https://doi.org/10.5281/zenodo.17852914 |