Modelling of the Output Power from a Grid-Connected Photovoltaic System Using the Artificial Neural Network

  • Shahril Irwan Sulaiman
  • Sulaiman Shaari
  • Ahmad Maliki Omar
Conference paper


This paper presents the modelling of the output power of grid-connected photovoltaic (GCPV) system using the multilayer feedforward neural network (MLFNN). Different combinations of solar irradiance, ambient temperature and module temperature were derived for determining the best type of parameters which could be used to model the output power using the MLFNN. The MLFNN was implemented in two stages, i.e. training and testing, and all data were obtained from a retrofitted-based GCPV system. The results show that the MLFNN model with solar irradiance and ambient temperature had outperformed the model with solar irradiance and module temperature as well as the model with solar irradiance, ambient temperature and module temperature in producing higher modelling accuracy.


Root Mean Square Error Artificial Neural Network Hide Layer Output Layer Solar Irradiance 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    Tao C, Shanxu D, Changsong C (2010) Forecasting power output for grid-connected photovoltaic power system without using solar radiation measurement. In: 2nd International symposium on power electronics for distributed generation systems, HeFei, China, pp 773–777Google Scholar
  2. 2.
    Mellit A, Pavan AM (2010) Performance prediction of 20kWp grid-connected photovoltaic plant at Trieste (Italy) using artificial neural network. Energ Convers Manag 51:2431–2441CrossRefGoogle Scholar
  3. 3.
    Skoplaki E, Palyvos JA (2009) Operating temperature of photovoltaic modules: a survey of pertinent correlations. Renew Energ 34:23–29CrossRefGoogle Scholar
  4. 4.
    Gao DW, Wand P, Liang H (2007) Optimization of hidden nodes and training times in ANN-QSAR model. Environ Informat Arch 5:464–468Google Scholar
  5. 5.
    Rao KVS, Krishna MR, Babu DB (2009) Cryptanalysis of a feistel type block cipher by feedforward neural network using right sigmoidal analysis. Int J Soft Comput 4:131–135Google Scholar
  6. 6.
    Souza AP, Escobedo JF (2013) Estimates of hourly diffuse radiation on tilted surfaces in Southeast of Brazil. Int J Renew Energ Res 3:207–221Google Scholar
  7. 7.
    Zainuddin H, Shaari S, Omar AM, Sulaiman SI, Mahmud Z, Muhamad Darus F (2011) Prediction of module operating temperatures for free-standing (FS) photovoltaic (PV) system in Malaysia. Int Rev Model Simulat 4:3388–3394Google Scholar
  8. 8.
    Arafa M, Alqedra M (2011) Early stage cost estimation of buildings construction projects using artificial neural networks. J Artif Intell 1:63–75Google Scholar
  9. 9.
    Moghaddam MG, Ahmad F, Basri M, Abdul Rahman MB (2010) Artificial neural network modeling studies to predict the yield of enzymatic synthesis of betulinic acid ester. Electron J Biotechnol 13:1–12Google Scholar

Copyright information

© Springer Japan 2015

Authors and Affiliations

  • Shahril Irwan Sulaiman
    • 1
  • Sulaiman Shaari
    • 2
  • Ahmad Maliki Omar
    • 1
  1. 1.Faculty of Electrical EngineeringUniversiti Teknologi MARAShah AlamMalaysia
  2. 2.Faculty of Applied SciencesUniversiti Teknologi MARAShah AlamMalaysia

Personalised recommendations