Prediction of solar energy and radiation in Singapore

• Main Author: Lee, Jian Wei.
• Other Authors: Chan Chee Keong
• Format: Final Year Project
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Power electronics
• Online Access: http://hdl.handle.net/10356/40021
• Institution: Nanyang Technological University

This project was to perform prediction of solar energy and radiation in Singapore. Artificial Neural Network (ANN) was used in this project for prediction purpose. Data obtained from National Weather Study Project (NWSP) were used as data sources. Java Object Oriented Neural Engine (JOONE) was used to run the network designed. In this report, the data used were in between 9am to 10am, all together there were 664 datasets. 564 datasets were used for training and the 100 datasets were used for validating the output of network. The error was evaluated by using the 100 validating datasets and their respective network outputs. The accuracies were evaluated within 5% error. In this project, we used the training datasets to teach the ANN so that the network could give us the desired output. The learning algorithm selected in this project was BackPropagation (BP). BP was a method using the difference between the target output value and the network output value to modify the weights of the network. The changes of weights were accumulated during the network learning process. After the learning process, the network was tested with validating data so as to evaluate the accuracy of the network.