An on-Line Fault Diagnosis System with Power Generation Promotion for Photovoltaic Module Arrays

• Main Authors: Meng-Cheng Wu, 吳孟承
• Other Authors: Kuei-Hsiang Chao
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/3596r3

碩士 === 國立勤益科技大學 === 電機工程系 === 105 === The main purpose of this thesis is to develop an on-line diagnosis system with power generation promotion for photovoltaic module arrays. The on-line diagnosis system of the proposed photovoltaic module arrays can enhance the photovoltaic power generation by connecting the modules in the form of total-cross-tied (TCT) when the photovoltaic modules are under partial shading or malfunctioning. In addition, the system can immediately find out the location of shading or fault module by line-current direction of crossing the modules for module troubleshooting and then to maintain the optimal photovoltaic power generation. Considering the fact that partial shading modules will cause the PV characteristic curve to have a multi-peak phenomenon, the maximum power point tracking method of improved teaching and learning based optimization (I-TLBO) will be used to track the global maximum power to ensure the module arrays can output the maximum power. In this thesis, researching the feasibility and adaptability of the teaching and learning algorithm for the maximum power tracking of photovoltaic module arrays, and attempts to improve teaching and learning algorithm by adding intelligent tracking and learning strategies, in order to enhance the tracking and response speed. And then, through the analysis of photovoltaic module arrays shading or failure of the characteristics phenomenon, and then it was used to determine the correct location of shading or fault module, and finally the information through the LCD to inform the maintenance staff for troubleshooting. In order to make the algorithm of the system have a good computing speed, the digital signal processor (DSP) will be used to implement the improved teaching and learning based optimization in this thesis, and complete an on-line diagnosis system for photovoltaic module arrays under partial shading or malfunctioning.