Assessing natural ventilation rates using a combined measuring and modelling approach

Vertaalde titel van de bijdrage: Bepalen van debiet in natuurlijk geventileerde stallen door gebruik te maken van meettechnieken en neurale netwerken

Gerlinde De Vogeleer, Philippe Van Overbeke, Jan G. Pieters, Peter Demeyer

    Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureC1: Artikels in proceedings van wetenschappelijke congressen, die niet inbegrepen zijn in A1, A2, A3 of P1peer review

    66 Downloads (Pure)

    Uittreksel

    Natural ventilation of animal houses clearly has advantages as for instance its low power consumption. However its application is often limited due to the lack of a reliable measuring and control system of the ventilation rate and so of emissions, as required for legislation.
    Although a lot of models exist to determine natural ventilation rates in buildings, it is still a challenge to know the ventilation rate accurately with few measurements. The objective of this work was to develop a model for the prediction of the natural ventilation rate in a pig house with as few measuring points as possible. Neural networks were used to investigate the reliability and accuracy of using as limited input as possible, taken from data collected from measurements with sonic anemometers in a real scale test building under outside weather conditions.
    Vertaalde titel van de bijdrageBepalen van debiet in natuurlijk geventileerde stallen door gebruik te maken van meettechnieken en neurale netwerken
    Oorspronkelijke taalEngels
    TitelInternational Conference of Agricultural Engineering - AgEng 2014 Zurich - Engineering for Improving Resource Efficiency
    Aantal pagina’s8
    UitgeverijThe European Society of Agricultural Engineers (EurAgEng)
    Publicatiedatum7-jul.-2014
    ISBN van elektronische versie978-0-9930236-0-6
    PublicatiestatusGepubliceerd - 7-jul.-2014
    EvenementAgEng 2014 - Zurich, Zwitserland
    Duur: 7-jul.-201410-jul.-2014
    http://www.ageng2014.ch

    Dit citeren