Performance evaluation of a split air conditioning system with artificial neural network approach

Bilgili M., Şimşek E., Çoşgun A., Yaşar A.

Energy Education Science and Technology Part A: Energy Science and Research, cilt.30, sa.SPEC .ISS.1, ss.83-94, 2012 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 30 Sayı: SPEC .ISS.1
  • Basım Tarihi: 2012
  • Dergi Adı: Energy Education Science and Technology Part A: Energy Science and Research
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED)
  • Sayfa Sayıları: ss.83-94
  • Anahtar Kelimeler: Artificial neural network, Coefficient of the performance, Refrigeration, Split air conditioning
  • Akdeniz Üniversitesi Adresli: Evet

Özet

This paper deals with predicting the performance of a split air conditioning (SAC) system using artificial neural network (ANN) approach. For this aim, an experimental R-22 split air conditioning system was developed and equipped with instruments used for temperature, pressure, current and power measurements. The experimental system was operated at steady state conditions varying the condenser inlet air temperature. Using some of the experimental data for training, an ANN model for the SAC system was developed. Inputs of the ANN model include the condenser inlet air temperature and evaporating temperature. Outputs of the ANN model consist of the ideal cooling and heating coefficients of performance, cooling and heating coefficients of performance, compressor isentropic efficiency, compressor power, cooling capacity, heat rejection rate in the condenser, refrigerant mass flow rate, evaporator inlet and outlet air temperatures, condenser outlet air temperature, condensing temperature and compressor current. The ANN predictions for these parameters usually agreed well with the experimental values with mean relative errors (MREs) in the range of 0.03-4.55%, root mean square errors (RMSEs) in the range of 0.0071-0.7573, and absolute fraction of variance (R2) in the range of 0.99798-1.00. This study shows that SAC systems can be alternatively be modeled using ANNs with a high degree of accuracy. © Sila Science.