Material Distribution Path Planning Based on Improved Parrot Optimization Algorithm

The complex and dynamic terrain of pumped storage power station construction sites presents significant challenges for efficient material delivery. Traditional path planning methods often struggle with slow convergence, poor global search ability, and local optima entrapment. To address these issues...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2025 5th International Conference on Machine Learning and Intelligent Systems Engineering (MLISE) s. 141 - 144
Hlavní autoři: Zhong, Jianxu, Li, Ritang, Zhou, Yupeng, Zhao, Bushi, Zhu, Junwei, Cui, Weihong, Yu, Shaofeng, Liao, Chongyang
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 13.06.2025
Témata:
Attention mechanism
Attention mechanisms
Chaotic mapping
Convergence
Heuristic algorithms
Improved Parrot Optimization Algorithm
Lenses
Machine learning algorithms
Optimization
Path planning
Power generation
Reverse learning
Simulation
Time-window constraints
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The complex and dynamic terrain of pumped storage power station construction sites presents significant challenges for efficient material delivery. Traditional path planning methods often struggle with slow convergence, poor global search ability, and local optima entrapment. To address these issues, this paper proposes an Improved Parrot Optimization Algorithm (IPOA) for material distribution routing. IPOA integrates chaotic mapping for better population initialization, an attention mechanism for adaptive weight adjustment, and a lens imaging-based reverse learning strategy to enhance exploration. A time-window constraint model is also introduced to reflect varying delivery urgency. Simulation results show that IPOA outperforms existing algorithms in convergence speed, adaptability, and solution quality, improving delivery efficiency and supporting project schedule adherence.
DOI:10.1109/MLISE66443.2025.11100260