Research on Establishing Forecast Model of PV Power Generating System by Neural Network Algorism
| Autor: | Yung-Kun, 楊騉 |
|---|---|
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 104 With the developed technology and popular application,PV Power Generating System is mostly used. For example, it is used in electricity in isolated island, aids to navigation, base station of cell phone and marked instruction on the load. However, the power generation of the photovoltaic is easily influenced by the sunlight climate. In this study, the ultraviolet index and other information of the climate is chosen from the Central Weather Bureau to predict the daily power generation. This is used to a Few Days Ahead Generating Power Forecasting. In this study, the back-propagation neural network is used to build the predicting model. The two predicting model is in the following. One is the solar PV system in National Taipei University of Technology. The other is the insolation of Taipei City. In the model, the ultraviolet index, the max temperature of the day, rain, other information of the weather forecast and the number of days to Summer Solstice are the inputs to predict the daily power generation and insolation of the second day. The experiment is to analyze the characteristic of the input and adjust the number of the cell. And it makes the model to learn moderately. The model is built to compare the actual information and between the models. The reaction of inputting the assumed information is also discussed. The model build in this study has the reasonable reaction and the prediction. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|