局部阴影下的光伏阵列MPPT算法研究

发布时间：2018-08-27 14:55

【摘要】：能源是人类社会发展的物质基础和动力来源。一方面世界经济的高速发展以及世界人口的持续增长,使得能源的总需求量正以惊人的速度增长；另一方面全球的化石燃料储量日益减少,并存在枯竭的危机。此外,化石燃料的大量使用造成了严重的环境问题。所以开发利用绿色可再生能源迫在眉睫。光伏发电是新兴的可再生能源技术,以其自身特点得到了人们的关注。本文以光伏发电系统作为研究对象,以获得遮阴下的光伏阵列的最大输出功率为研究目标。从以下几个方面做了具体的研究。 1)基于matlab simulink搭建了光伏电池以及Boost电路的仿真模型。分析了光伏电池以及遮阴光伏阵列的输出特性曲线的变化规律,为遮阴光伏阵列的最大功率点跟踪(MPPT)算法设计做准备。实测了两种天气下的光伏电池的峰值点数据,实测数据具有的规律与仿真分析的结果具有较好的一致性。 2)对比分析了四种传统的MPPT算法,包括恒定电压法、扰动观察法、增量电导法以及模糊逻辑控制法。在算法分析的过程中,为后三种算法设计了不同的变步长控制策略,并对它们各自的控制效果作了仿真分析。本文介绍了一种基于人工神经网络通过实测数据生成模糊逻辑控制规则的方法,为模糊控制规则的在线生成提供了条件。 3)存在局部阴影时,光伏阵列的功率-电压特性(P-V)曲线出现多个极值点、电流-电压(I-V)特性曲线呈现阶梯状,这使得基于单峰寻优的传统MPPT算法失效。为此,本文在研究遮阴光伏阵列输出特性规律的基础上,提出了一种具有全局搜索能力的MPPT算法。该算法先用粒子群优化(PSO)算法将输入位置调整到全局最优附近,再用变步长电导增量法得到全局最优解。新的算法在减轻系统振荡和加快搜索速度方面做了改进。仿真结果表明,该方法不仅较好地克服了现有算法使用PSO大幅度随机初始化粒子位置而导致的系统振荡问题,而且有效利用传统单峰寻优算法的优点,增强了系统搜索的快速性和稳定性,取得了较好的控制效果。 4)本文最后对小型光伏发电系统的硬件电路进行了设计。采用软开关技术实现的Boost电路提高了功率电路的转换效率。
[Abstract]:On the one hand, the rapid development of the world economy and the sustained growth of the world population make the total demand for energy increasing at an alarming rate; on the other hand, the global fossil fuel reserves are decreasing day by day, and there is a crisis of exhaustion. It has become a serious environmental problem. So the development and utilization of green renewable energy is imminent. Photovoltaic power generation is a new renewable energy technology, which has attracted people's attention for its own characteristics. Specific aspects have been studied.
1) The simulation models of photovoltaic cells and Boost circuits are built based on MATLAB simulink. The output characteristic curves of photovoltaic cells and shaded photovoltaic arrays are analyzed to prepare for the design of maximum power point tracking (MPPT) algorithm for shaded photovoltaic arrays. According to the laws and the results of simulation analysis, the results are in good agreement.
2) Four traditional MPPT algorithms are compared and analyzed, including constant voltage method, disturbance observation method, incremental conductance method and fuzzy logic control method. In the process of algorithm analysis, different variable step size control strategies are designed for the latter three algorithms, and their respective control effects are simulated and analyzed. The method of generating fuzzy logic control rules from measured data by network provides conditions for on-line generation of fuzzy control rules.
3) The power-voltage characteristic (P-V) curves of photovoltaic arrays with local shadows have multiple extremum points, and the current-voltage (I-V) characteristic curves are ladder-shaped, which makes the traditional MPPT algorithm based on single-peak optimization invalid. MPPT algorithm of force. The algorithm adjusts the input position near the global optimum by particle swarm optimization (PSO) algorithm, and then obtains the global optimum by variable step conductance increment method. The problem of system oscillation caused by large-scale random initialization of particle position is solved, and the advantages of traditional single-peak optimization algorithm are effectively utilized to enhance the rapidity and stability of system search, and good control effect is achieved.
4) At last, the hardware circuit of the small-scale photovoltaic power generation system is designed. The Boost circuit realized by soft-switching technology improves the conversion efficiency of the power circuit.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM615;TP18

【参考文献】