A hybrid multi-start metaheuristic scheduler for astronomical observations

In this paper, we investigate Astronomical Observations Scheduling which is a type of Multi-Objective Combinatorial Optimization Problem, and detail its specific challenges and requirements and propose the Hybrid Accumulative Planner (HAP), a hybrid multi-start metaheuristic scheduler able to adapt...

Descripción completa

Detalles Bibliográficos
Autores: Nakhjiri, Nariman, Salamó Llorente, Maria, Sànchez i Marrè, Miquel, 1964-, Morales, Juan Carlos
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2023
País:España
Institución:Varias* (Consorci de Biblioteques Universitáries de Catalunya, Centre de Serveis Científics i Acadèmics de Catalunya)
Repositorio:Recercat. Dipósit de la Recerca de Catalunya
OAI Identifier:oai:recercat.cat:2445/218521
Acceso en línea:https://hdl.handle.net/2445/218521
Access Level:acceso abierto
Palabra clave:Intel·ligència artificial
Observacions astronòmiques
Aprenentatge automàtic
Artificial intelligence
Astronomical observations
Machine learning
Descripción
Sumario:In this paper, we investigate Astronomical Observations Scheduling which is a type of Multi-Objective Combinatorial Optimization Problem, and detail its specific challenges and requirements and propose the Hybrid Accumulative Planner (HAP), a hybrid multi-start metaheuristic scheduler able to adapt to the different variations and demands of the problem. To illustrate the capabilities of the proposal in a real-world scenario, HAP is tested on the Atmospheric Remote-sensing Infrared Exoplanet Large-survey (Ariel) mission of the European Space Agency (ESA), and compared with other studies on this subject including an Evolutionary Algorithm (EA) approach. The results show that the proposal outperforms the other methods in the evaluation and achieves better scientific goals than its peers. The consistency of HAP in obtaining better results on the available datasets for Ariel, with various sizes and constraints, demonstrates its competence in scalability and adaptability to different conditions of the problem.