• Mi UCrea
    Ver ítem 
    •   UCrea
    • UCrea Académico
    • Facultad de Ciencias
    • Máster Universitario en Matemáticas y Computación
    • M1098 Trabajos académicos
    • Ver ítem
    •   UCrea
    • UCrea Académico
    • Facultad de Ciencias
    • Máster Universitario en Matemáticas y Computación
    • M1098 Trabajos académicos
    • Ver ítem
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    A model to predict water consumption in growing pigs

    Ver/Abrir
    MarrocoRiosMaria.pdf (14.51Mb)
    Identificadores
    URI: https://hdl.handle.net/10902/34902
    Compartir
    RefworksMendeleyBibtexBase
    Estadísticas
    Ver Estadísticas
    Google Scholar
    Registro completo
    Mostrar el registro completo DC
    Autoría
    Marroco Ríos, María
    Fecha
    2024-09
    Director/es
    Palazuelos Calderón, CamiloAutoridad Unican
    Derechos
    © María Marroco Ríos
    Resumen/Abstract
    Water consumption plays a vital role in pig farming operations, and accurate prediction of water use is essential for efficient resource management. This study, part of the European DECIDE project—a five-year initiative aimed at developing data-driven decision support tools that provide robust and early signals of disease onset and options for diagnostic confirmation—aims to develop a comprehensive model to predict water consumption in pig herds, considering various external variables and incorporating historical water consumption data from previous batches. First, the key parameters influencing water consumption in growing pigs, such as feed intake, humidity, indoor farm temperature, and days of growth, are identified, a task not done before. These factors are integrated into the predictive model, which was developed using both R and Python programming language. To capture the daily seasonal pattern in water consumption, four harmonics were incorporated into the model. Through experimentation, it was discovered that adding a fourth harmonic improved prediction accuracy. Additionally, the model was enhanced by incorporating a parameter that included historical water consumption data from other batches produced on the same day and at the same growth stage. This additional feature proved effective in improving the model’s performance. Beyond creating this model and capturing the upward trend in water consumption of a batch and daily seasonality, we also determined the daily water consumption patterns on a warm summer day and a cold winter day, noting that the pattern varies depending on the season. The developed model was validated using historical data from several batches of pigs on the same farm. The predicted water consumption values were compared with actual values, and it was observed that all actual values were within the confidence intervals calculated using the Delta method, validating the model’s reliability and accuracy. Thus, this study successfully achieved the objective of modelling water consumption in pig herds. The findings of this research contribute to the efficient management of water resources in pig farming and lay the groundwork for future advancements in this field.
    Colecciones a las que pertenece
    • M1098 Trabajos académicos [53]

    UNIVERSIDAD DE CANTABRIA

    Repositorio realizado por la Biblioteca Universitaria utilizando DSpace software
    Contacto | Sugerencias
    Metadatos sujetos a:licencia de Creative Commons Reconocimiento 4.0 España
     

     

    Listar

    Todo UCreaComunidades y coleccionesFecha de publicaciónAutoresTítulosTemasEsta colecciónFecha de publicaciónAutoresTítulosTemas

    Mi cuenta

    AccederRegistrar

    Estadísticas

    Ver Estadísticas
    Sobre UCrea
    Qué es UcreaGuía de autoarchivoArchivar tesisAcceso abiertoGuía de derechos de autorPolítica institucional
    Piensa en abierto
    Piensa en abierto
    Compartir

    UNIVERSIDAD DE CANTABRIA

    Repositorio realizado por la Biblioteca Universitaria utilizando DSpace software
    Contacto | Sugerencias
    Metadatos sujetos a:licencia de Creative Commons Reconocimiento 4.0 España