光伏并网发电系统中基于状态估计的MPPT及孤岛检测方法研究

发布时间：2019-06-18 18:09

【摘要】：局部阴影的情况下的光伏阵列的最大功率点,具备了多个局部功率峰值和时变不确定以及非线性的特点。针对这些特点,本文应用现代状态估计理论,建成了光伏阵列输出功率动态模型,同时利用了交互多模型估计的算法,比较准确地跟踪了多峰功率多状态曲线,因而抑制了噪声,定位出最大的功率点,最后,提出了基于多模型状态估计的光伏阵列最大功率点跟踪(Maximum Power Point Tracking,MPPT)技术。实验及仿真验证所提出的新MPPT控制策略其有效性以及正确性的特点,在有无阴影的情形下均能准确和快速地跟踪最大功率点,进而提高了光伏发电的效率。光伏发电系统多逆变器并网是本文的研究对象,在孤岛效应检测中引入了现代状态估计理论,同时提出一种新的方法,即基于PCC电压正弦状态估计的被动孤岛检测方法,推导出电压正弦波状态估计模型并离散化,利用强跟踪滤波器,随后对此电压状态进行实时监测。通过BP神经网络对实时状态参数变化进行识别,因此可对伪孤岛、孤岛以及正常情况下的不同现象进行区分辨别。通过对仿真及实验结果的分析,证明了此方法能够快速且有效地将孤岛检测出来。
[Abstract]:The maximum power point of photovoltaic array in the case of local shadow has the characteristics of multiple local power peaks, time-varying uncertainties and nonlinear characteristics. In view of these characteristics, in this paper, the dynamic model of photovoltaic array output power is established by using modern state estimation theory, and the interactive multi-model estimation algorithm is used to track the multi-peak power multi-state curve more accurately, so as to suppress the noise and locate the maximum power point. Finally, the maximum power point tracking (Maximum Power Point Tracking,MPPT technology of photovoltaic array based on multi-model state estimation is proposed. Experiments and simulations verify the effectiveness and correctness of the proposed new MPPT control strategy, and can track the maximum power point accurately and quickly with or without shadow, thus improving the efficiency of photovoltaic power generation. Multi-inverter grid-connected in photovoltaic power generation system is the research object of this paper. The modern state estimation theory is introduced into the islanding effect detection. At the same time, a new method is proposed, that is, the passive islanding detection method based on PCC voltage sinusoidal state estimation. The voltage sine wave state estimation model is deduced and discretized, and then the voltage state is monitored in real time by using a strong tracking filter. The real-time state parameters can be identified by BP neural network, so the pseudo-isolated islands, isolated islands and different phenomena under normal conditions can be distinguished and distinguished. Through the analysis of simulation and experimental results, it is proved that this method can detect isolated islands quickly and effectively.
【学位授予单位】：南昌大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM615

【参考文献】