Performance evaluation of solar thermal systems using an artificial neural network | |
| Author | Somjet Leeludej |
| Call Number | AIT Thesis no. ET-01-20 |
| Subject(s) | Solar thermal energy Neural networks (Computer science) |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, 2001 |
| Publisher | Asian Institute of Technology |
| Abstract | Solar thermal systems are usually evaluated by experimental and/or theoretical simulation studies. The experimental method is a foolproof method, but is time consuming and maybe expensive. Theoretical simulation studies help to evaluate systems even before they are built. TRNSYS is a powetful design tool, which allows detailed analysis of all components of the system. Artificial Neural Networks (ANN) are also powelful tools to help predict the pe1formance of energy systems. The aim of this study was to conduct pe1formance evaluation studies of solar thermal systems (water and air heating) using ANN and compare the pelformance of solar thermal systems between experimental studies and simulation (TRNSYS and ANN). Experimental studies on solar thermal systems were conducted for a forced circulation solar water heating system (14 days of experiments), a forced circulation solar air heating system (16 days of experiments), and a number of domestic water heating system (natural circulation) in Thailand (9 systems). ANN studies included training to predict the water temperature at collector outlet, mean water temperature in storage tank, and useful energy delivered from the collector of forced circulation solar water heating system (at hourly intervals), air temperature at collector outlet and useful energy delivered from the collector of solar air heating system (at hourly intervals), and outlet temperature of water withdrawn of the system and useful energy delivered from the system of domestic solar water system (daily values). The statistical parameters and the maximum deviation were used to measure the accuracy of the prediction approaches. The pelformances of solar thermal system were predicted for 3 different climatic conditions (clear, partly cloudy, and cloudy day) for forced circulation solar water and air heating systems A comparison of the pelformance prediction between simulation studies (TRNSYS and ANN) and experiments of solar thermal systems, shows that both the simulation methods gave very good prediction and TRNSYS always gave better accurate results than ANN. The prediction by ANN could be improved by using more data sets covering the wide range of climatic conditions. |
| Year | 2001 |
| Type | Thesis |
| School | School of Environment, Resources, and Development (SERD) |
| Department | Department of Energy and Climate Change (Former title: Department of Energy, Environment, and Climate Change (DEECC)) |
| Academic Program/FoS | Energy Technology (ET) |
| Chairperson(s) | Kumar, S.; |
| Examination Committee(s) | Surapong Chirarattananon ;Raj apakse, A.D.; |
| Scholarship Donor(s) | Royal Thai Government (RTG) ; |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 2001 |