户用光伏微网系统研究

发布时间：2018-03-23 23:16

本文选题：光伏微网　切入点：遗传算法　出处：《山东农业大学》2014年博士论文

【摘要】：由于户用型光伏微网位于电网终端,负载能就地消耗光伏电池所发出电能,从而使电能在传输过程中的损失大大减少,充分利用了分布式光伏能源发电紧靠本地负荷中心和无污染的优点,同时又克服了太阳能地域分散性的缺点。和光伏大电站相比,户用光伏微电网具有灵活性和经济性等优点。因此,本文以光伏发电与储能系统连接到电网的基础上,分析了户用光伏微网的拓扑结构和孤岛检测方法研究以及户用光伏微网的计划孤岛运行,并且户用光伏微网在孤岛运行时,如何使逆变器输出的电压和频率以及波形的畸变率符合电能质量的要求,以及利用遗传算法分析并制定了了该户用光伏微网系统的日前调度计划。本文的主要工作和成果有:1、设计了一台3 kW的单相光伏并网逆变器,既能并网运行,又能离网运行,输出电压为220 V,50 Hz的正弦波,额定功率为3 kW,波形畸变率小于5%,功率因数等于1,符合设计要求。2、设计了一套户用光伏发电微电网装置。在对光伏并网逆变器分析研究的基础上,设计了一套户用光伏微电网装置,包括3kW光伏电池板、两组12 V、200 Ah蓄电池、光伏并网逆变器3 kW、双向逆变器3kW和本地交流负荷等,通过获取的户用光伏微网数据研究从而验证光伏微电网平滑切换的主从控制策略等。理论分析、仿真及实验数据表明:提出的户用光伏微电网拓扑结构,既能并网运行,又能离网运行,所提出的主从控制策略能高效的吸收光伏电能,高效率地为本地负荷提供电能,确保系统稳定可靠运行。3、为了智能电网发展的需要,本文以山东泰安地区某户负荷为例,结合基于储能水平控制,采用遗传算法合理制定户用光伏微电网日前用电计划。得到结论如下:(1)用户角度该户用光伏微电网优化前可售电情况下从电网售电能获得收入2.76元,所建立的模型能够通过合理安排储能水平和可时移负荷的用电时间,使优化后可售电情况下能从电网获得4.925元,从而达到家庭收入最多,光伏发电能源的发电利用率最大。(2)电网角度对可时移负荷进行优化调度后,通过储能和可时移负荷的配合可以起到削峰填谷的作用,对增加供电能力和提高供电可靠性有重要意义。4、针对被动式孤岛检测方法有很大的非检测区和主动式孤岛检测方法对电能质量有谐波污染等缺点,本文对并网逆变器提出了一种新的被动式孤岛检测方法—基于小波神经网络判别孤岛检测。该方法采集公共连接点的电压信号,利用基于各尺度小波变换系数能量的提取特征量方法,提取出的特征向量不仅具有平移不变的特性,还反映信号的时频局部特征,提高了孤岛检测的实时性。采用具有强大的模式识别能力的三层BP神经网络,能有效辨识孤岛和非孤岛状态。该孤岛检测方法通过Matlab仿真结果表明判别孤岛与非孤岛的正确率达到了98.45%,比传统的被动式孤岛检测方法检测盲区小,由于没有向控制信号中加入扰动量,因而没有谐波污染,不会对电能质量产生不良影响,这弥补了主动式检测法的不足,并且不会引起电压幅值或频率越限,对系统没有破坏性,也不会带来有功功率波动,有利于光伏能源应用。
[Abstract]:Because of household photovoltaic micro grid in power terminal, can load local consumption of photovoltaic cells emitted by power, greatly reduced so that the electric power transmission in the process of loss, make full use of the distributed photovoltaic power sources are close to the load center and has no pollution, but also overcome the shortcomings of solar geographicdispersion compared. And the photovoltaic power station, PV micro grid has the advantages of flexibility and economy. Therefore, the photovoltaic power generation and energy storage system connected to the grid on the basis of the analysis of topological structure and islanding detection method for photovoltaic microgrid households and households with photovoltaic micro plan island operation network, and household photovoltaic microgrid in islanding operation, how to make the inverter output voltage and frequency and waveform distortion rate to meet the requirements of power quality, and the use of genetic algorithm is analyzed and formulated by the user For photovoltaic microgrid system before the date of dispatch plan. The main work and achievements are: 1, the design of the single phase photovoltaic grid connected inverter with a 3 kW, and can operate in parallel to the grid, off grid, output voltage of 220 V and 50 Hz sine wave, the rated power is 3 kW, waveform distortion rate less than 5%, the power factor is equal to 1, meet the design requirements of.2, a set of household photovoltaic micro grid device design. Based on the analysis of photovoltaic inverter, designed a set of micro grid photovoltaic device for users, including 3kW photovoltaic panels, two groups of 12 V, 200 Ah batteries, photovoltaic the grid connected inverter 3 kW bi-directional inverter 3kW and local AC load, through the acquisition of the PV micro research network data to verify the photovoltaic microgrid smooth switching master-slave control strategy. Theoretical analysis, simulation and experimental data show that the proposed PV micro grid topology, Can network operation, and can run from the network, the master-slave control strategy proposed can absorb the photovoltaic power efficient, high efficiency for the local load power, to ensure stable and reliable operation of.3 system, in order to satisfy the need of the development of smart grid, this paper takes Shandong Tai'an area of a household load as an example, combined with the energy level control based on genetic algorithm is used to establish the PV micro electricity grid before the plan. The conclusions are as follows: (1) the user angle of the household photovoltaic micro grid optimization before the sale of electricity from the sale of electricity grid can obtain the income of 2.76 yuan, the model can through the reasonable arrangement and storage level when the load of the power of the time, so that after optimization, the sale of electricity can get 4.925 yuan from the grid, so as to achieve the maximum family income, photovoltaic energy generation utilization. (2) when the load angle on the grid can be optimized after scheduling Through the cooperation, energy storage and time shift load can play a role in load shifting, to increase the power supply capacity and.4 have important significance to improve the reliability of power supply, the passive islanding detection methods have great non detection zone and active islanding detection method of power quality harmonic pollution and other shortcomings, is presented in this paper. A new passive islanding detection method based on wavelet neural network discriminant islanding detection of grid connected inverter. This method collects public connection point of the voltage signal, using the method of feature extraction of wavelet transform coefficients of each scale of the amount of energy based on the extracted feature vector not only has the characteristics of translation invariance, but also reflect the local time-frequency characteristics of the signal and improve the real-time of islanding detection. By using the three layer BP neural network has a strong ability to pattern recognition, can effectively identify islanding and non islanding state. The islanding detection. The Matlab simulation results show that the correct rate of discrimination and non isolated island reached 98.45%, than the traditional method of passive islanding detection zone is small, because there is no disturbance to join the control signal, so there is no harmonic pollution on the power quality, will not have a negative impact, which makes up the deficiency of active detection method. And will not cause the voltage amplitude or frequency limit, the system is not destructive, will not lead to the fluctuation of active power, is conducive to solar energy application.

【学位授予单位】：山东农业大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM615

【参考文献】