Titelaufnahme

Titel
Implementation of simulation-based virtual sensors using Radiance / von Vaibhav Jain
VerfasserJain, Vaibhav
Begutachter / BegutachterinMahdavi, Ardeshir ; Zach, Robert
Erschienen2014
UmfangV, 59 Bl. : Ill., zahlr. graph. Darst.
HochschulschriftWien, Techn. Univ., Mag.-Arb., 2014
SpracheEnglisch
DokumenttypDiplomarbeit
Schlagwörter (EN)Simulation-based service / Virtual sensors / Radiance / Software development / JAVA
URNurn:nbn:at:at-ubtuw:1-74381 Persistent Identifier (URN)
Zugriffsbeschränkung
 Das Werk ist frei verfügbar
Dateien
Implementation of simulation-based virtual sensors using Radiance [7.18 mb]
Links
Nachweis
Klassifikation
Zusammenfassung (Englisch)

Building performance simulation is traditionally used to support the building design process; however, they can also be used during building operation phase by providing simulation output as control parameter for building control and/or monitoring systems. Such a simulation output from a virtual (or computational) model of a building are hereafter referred as output from virtual sensor. Theoretically, virtual sensors, if reliably and effectively incorporated, can expand the reach of real sensors (Zach, Glawischnig, & Mahdavi). A continuous supply of output from such sensors can support real-time (or near real-time) monitoring of simple environmental conditions such as air temperature, relative humidity etc. as well as complex performance indicators such as thermal comfort, visual glare etc. This information, especially in big multi-zone buildings, will otherwise be very costly to obtain from real sensors due to high initial costs of the sensors, their installation, networking, and maintenance. Virtual sensors can easily increase the resolution of sensory information, since they have no monetary or spatial limitations. For example, a simulation model can have as many costly glare sensors which can be placed where ever desired. However, practically the accuracy of virtual sensors poses a challenge, especially in cases where high uncertainty is involved, such as daylight illuminance calculation. For accurate results (which mean being reasonably close to measured values from real sensors) well calibrated models of the building as well as its boundary conditions are required. This study will (a) describe the concept of virtual sensors, (b) implement this concept by developing a software application that automates the process of continuously producing real-time output from illuminance type virtual sensors, (c) create a calibrated model of a building situated in Vienna using Radiance lighting simulation program to test the developed software application, (d) discuss results, and challenges pertaining to the implementation. A comparison of measured and simulated output from the current version of developed application (Radiance based) was performed, and was found to give reliable results on clear sky days, but moderately accurate results on intermediate sunny/cloudy days. Results for artificial lighting case had good accuracy. The use of sky-scanner is recommended so that the simulation sky has the luminance distribution of the actual. Due to the modular architecture of the developed application, data from precise measurement instruments such as sky-scanner can be included in the future version of the application.