Skip to Main Content
Solar energy is one of the most promising renewable energy sources. In order to integrate this type of source into an existing power distribution system, system planners need an accurate model that predicts the availability of the generating capacity. Solar resources are known to exhibit a high variability in space and time due to the influence of other climatic factors such as cloud cover. The probability distribution of irradiance fluctuations is difficult to predict due to various uncertainties. For efficient conversion and utilization of the solar resource, the solar resource modelling is one of the most essential tools for proper development, planning, maintenance scheduling and pricing of solar energy system. This paper proposes the Mathematical and Neural Network Prediction models for estimation of solar radiation for Singapore. Meteorological and geographical data (latitude, longitude, altitude, month, mean sunshine duration, etc.) were used as inputs to the models. The estimated results are compared with the field data obtained from the pyranometer installed on the solar panel with a tilt of 15°. The relevance and performance of each model in Singapore's weather context is then evaluated using statistical tools, namely Mean Bias Error, Root Mean Squared Error and Mean Absolute Percentage Error. The results show that the correlation coefficients between the proposed model and the actual daily solar radiation were higher than 90%, thus suggesting a high reliability of the model for evaluation of solar radiation received in Singapore. These models can be used easily for estimation of solar radiation for preliminary design of solar applications.