Research On Intelligent Early - Warning System Of Intelligent Photovoltaic Grid - Connected System And Irradiation Fluctuation

With the drying up of fossil energy and the increasing of ecological environment pollution, the exploration, exploitation and utilization of renewable energy are imperative. Regarding with many green energy technologies, solar photovoltaic(PV) power generation with the merits of noiseless, regional limitless, easy expansion, longer service time and strong practicability has become one of the most important forms of solar energy utilization. The stability and security of PV grid power system is the important subject in the research on utilization technology of solar energy. The change of solar irradiance has a great impact on the PV grid power system, the solar irradiance affected by various types of random factors such as the weather, the geographical location and the clouds presents volatility and randomness. These factors are not conducive to the stable operation of grid systems, and produce the adverse influence on the power quality output. Exploring and finding a precise collection and forecasting method of solar irradiance information, not only can ensure the operation reliability and stability of PV grid power system, but also can output better electricity, which can provide the strong support for the optimization of system performance.The intelligent PV grid power system based on power line carrier(PLC) communication is introduced in this paper, and compared the performance with the traditional PV grid system. The experimental results show that when the partial shadow shade proportion of PV modules is between 0% and 25%, the biggest mismatch loss of traditional PV grid power system is about 35% due to the change of partial shadow shade. When the inclination of PV modules change between 0° and 90°, the biggest mismatch loss of traditional PV grid power system is about 5.8% due to the inclination changed. When the azimuth angle of PV modules change between 0° and 180°, the biggest mismatch loss of traditional PV grid power system is about 27.07% due to the azimuth changed.The solar irradiance warning system is proposed for PLC technology based intelligent PV grid-connected system. The low cost information collection instrument for solar irradiance is designed and the prediction algorithm of solar irradiance is proposed for the information collection.The monocrystalline silicon cell is selected in the thesis as the photoelectricity sensor based on the theoretical and experimental analysis. ATmega16 is using as the controller in the system. DC current collecting circuit, 16 bit analog-digital conversion circuit and the power module circuit of PLC are designed. By program the controller with high precision program for the DC current collecting, the low cost measuring device for solar irradiance is achieved. The function of data collection is verified.Combining the simple exponential smoothing with the kalman filtering method as the prediction algorithm and the exponential smoothing model through the change of solar irradiance data, the irradiance forecast for future’s 5 samples is realized. Utilizing the exponential smoothing prediction algorithm based on the state space model, the best state space model through a period of the history data changed of solar irradiance is selected. The change of future solar irradiationis predicted by utilizing the best state space model, and the forecast of solar irradiance for future 2 hours is done.