Optimal sizing of photovoltaic/battery/diesel based hybrid system and optimal tilting of solar array using the artificial intelligence for remote houses in India

• Published in: Energy and buildings Vol. 96; pp. 40 - 52
• Main Authors: Jeyaprabha, S. Berclin, Selvakumar, A. Immanuel
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.06.2015
• Subjects: Adaptive neuro fuzzy inference system; Artificial neural network; Hybrid PV/battery/diesel generator system sizing; LLP; Optimal tilt angle; LLP; Hybrid PV/battery/diesel generator system sizing; Adaptive neuro fuzzy inference system; Artificial neural network; Optimal tilt angle
• ISSN: 0378-7788
• DOI: 10.1016/j.enbuild.2015.03.012

•Optimal sizes of hybrid system and optimal tilt angle of solar array are performed for cities and predicted for the Indian remote houses with no grid access.•Hourly usage of diesel generator is reduced by optimal tilting.•Number of site visits for manual tracking are reduced to three per year for Indian locations.•Optimal sizing for different load demands is done through linear multiplication.•Hybrid system cost has been reduced below the standalone system cost for any load demand with loss of load probability lesser than 0.01. The optimal sizing and tilting of a hybrid photovoltaic/battery/diesel generator system are performed in this paper for the remote locations in India, using artificial intelligence techniques (AIT) without the metrological data. Initially, the optimal sizing and tilt angle calculation were done for different cities of India, for low cost and zero load rejection with the available metrological data. Using the latitude, longitude and altitude of any remote location, the optimal size of a hybrid system is found through AIT. The tilt angle for the photovoltaic (PV) array to be installed in any remote location is also predicted to reduce the hourly usage of diesel generator (DG) for all the four seasons and the number of visits for manual tracking is optimized to three per year through this research. The predicted optimal values, using adaptive neuro fuzzy inference system (ANFIS) and artificial neural network (ANN) are compared with the calculated values. The life cycle cost (LCC) of the optimized hybrid system is compared with the standalone PV as well as the DG system cost to prove its cost effectiveness. The validity of the sizing procedure for different load demand is also proved with loss of load probability (LLP) of 0.0026.