RT info:eu-repo/semantics/article
T1 Virtual Sensor Based on a Deep Learning Approach for Estimating Efficiency in Chillers
A1 Alonso Castro, Serafín
A1 Morán Palao, Antonio
A1 Pérez, Daniel
A1 Reguera Acevedo, Perfecto
A1 Díaz, Ignacio
A1 Domínguez, Manuel
A2 Ingenieria de Sistemas y Automatica
K1 Ingeniería de sistemas
K1 Virtual sensor
K1 HVAC systems
K1 Cooling power
K1 Eficiencia
K1 Deep Learning
K1 Convolutional Neural Network
K1 automática
AB Intensive use of heating, ventilation and air conditioning(HVAC) systems in buildings entails an analysis and monitoring of theire ciency. Cooling systems are key facilities in large buildings, and par-ticularly critical in hospitals, where chilled water production is neededas an auxiliary resource for a large number of devices. A chiller plantis often composed of several HVAC units running at the same time, be-ing impossible to assess the individual cooling production and e ciency,since a sensor is seldom installed due to the high cost. We propose avirtual sensor that provides an estimation of the cooling production,based on a deep learning architecture that features a 2D CNN (Convolu-tional Neural Network) to capture relevant features on two-way matrixarrangements of chiller data involving thermodynamic variables and therefrigeration circuits of the chiller unit. Our approach has been testedon an air-cooled chiller in the chiller plant at a hospital, and comparedto other state-of-the-art methods using 10-fold cross-validation. Our re-sults report the lowest errors among the tested methods and include acomparison of the true and estimated cooling production and e ciencyfor a period of several days
PB Springer
SN 1865-0929
LK http://hdl.handle.net/10612/11292
UL http://hdl.handle.net/10612/11292
NO International Conference on Engineering Applications of Neural Networks, 2017
NO P. 307-319
DS BULERIA. Repositorio Institucional de la Universidad de León
RD 25-abr-2024