COMPARISON BETWEEN NEURONAL AND EXPERIMENTAL CHARACTERIZATION OF INTEGRATED COLLECTOR-STORAGE SOLAR WATER HEATERS

摘要

Artificial Neural Networks (ANN) are widely accepted as a technology offering an alternative way to tackle complex and ill-defined problems. They have been used in diverse applications and have shown to be particularly effective in system identification and modeling as they are fault tolerant and can learn from examples. On the other hand, ANN are able to deal with non-linear problems and once trained can perform prediction at high speed. The objective of this work is the characterization of the integrated collector-storage solar water heater (ICSSWH) by the determination of the day time thermal (and optical) properties, and Night time heat loss coefficient with experimental temperatures, and predictive temperatures by (ANN). Because of that, an ANN has been trained using data for three types of systems, all employing the same collector panel under varying weather conditions. In this way the network was trained to accept and handle a number of unusual cases. The data presented as input were, the working systems (day or night), the type of system, the year, the month, the day, the time, the ambient air temperature, and the solar radiation. The network output is the temperature of the four tanks of storage unit. The correlations coefficients (R~2 -value) obtained for the training data set was equal to 0.997, 0.998, 0.998, and 0.996 for the four temperatures of each tank. The results obtained in this work indicate that the proposed method can successfully be used for the characterization of the ICSSWH.