Data Science and the Applications — Using Photovoltaic System Fault Detection for example

• Main Authors: CHEN, JIAN-RONG, 陳建融
• Other Authors: HSU, JIH-SHIH
• Format: Others
• Language: en_US
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/cavp8a

碩士 === 國立雲林科技大學 === 資訊管理系 === 104 === Data science is a popular issue nowadays and it can create enormous business value by using data appropriately. In Taiwan, solar energy is an emerging industry and encounters some challenges about photovoltaic (PV) plant maintenance. With the development of industry, the more PV plant is built the more demand of maintenance increase and this work is a heavy labor-oriented. Data science has some solution which can help solar energy industry to solve those challenges by implement fault detection. The main purpose of this research is applying data science technique to implement fault detection in solar energy industry. This research uses four different analysis techniques to achieve the purpose. The result discovers that Expectation-Maximization (EM), Density-Based Spatial Clustering of Applications with Noise (DBSCAN), and Ordering Points to Identify the Clustering Structure (OPTICS) can conduct photovoltaic fault detection. Each technique has a good outcome and can be applied to photovoltaic fault detection.