| Summary: | 碩士 === 國立勤益科技大學 === 電機工程系 === 98 === The main purpose of this thesis is to design and implement a portable fault diagnosis meter for a photovoltaic (PV) power generation system. The extension theory and extension neural network(ENN) theory are adopted to diagnose the fault type of a PV power generation system, respectively. A PV power generation system formed with 9 series and 2 parallels Sharp’s PV module NT-R5E3E is constructed by Solar Pro software for simulation. The simulation data is collected at different irradiances and temperatures of a PV power generation system under normal operation and various fault conditions. The gathered data is not only used to construct the classical domains and neighborhood domains of a extension theory, but also to construct the weights of a extension neural network theory. Some simulation results are made to demonstrate the effectiveness of the extension theory and extension neural network theory. The proposed fault diagnosis method of the PV power generation system needs immediate irradiances and temperatures of PV modules and the generation data of PV system, therefore a sensed circuit which can capture the fault characteristic data is designed and implemented. In this thesis, a PIC single chip was adopted to realize a portable photovoltaic power generation system fault diagnosis meter which is established with the extension theory, extension neural network theory and the ZigBee wireless network. Finally, experimental results are also carry out verify the accuracy of the proposed portable fault diagnosis meter for a PV power generation system.
|
|---|
