光伏发电系统MPPT控制算法的优化研究

发布时间：2018-06-01 17:36

本文选题：光伏电池 + 智能算法　；参考：《中国地质大学(北京)》2016年硕士论文

【摘要】：为了解决环境污染和能源短缺的压力,太阳能作为最环保、最便捷、可持续的新能源,拥有着巨大的开发潜能。光伏发电产业也备受关注,然而光伏系统的高投资成本和较低的转化效率成为影响光伏产业向前发展的主要因素。如何提高光伏发电的转换效率俨然已成为各国关注和讨论的焦点。因此,作为提高能效重要手段之一的光伏发电最大功率点跟踪技术应运而生。本文通过对光伏电池和各种控制算法的分析研究发现:在跟踪控制过程中,传统的一些控制算法很难实现动态响应速度和稳态精度的有效平衡,跟踪效率并不理想。本文创造性的提出将蛙跳粒子群神经网络算法应用到光伏最大功率点跟踪控制中来,利用智能算法强大的自学习能力、非线性拟合能力和容错能力等优势,在结合恒压控制思路,使跟踪效率大大提高。针对光伏电池的原理、特性,进行了详细的分析;在MATLAB中建立电池模型,仿真分析环境改变时对电池的U-I、U-P特性曲线的影响。为光伏控制系统建模提供基础。剖析MPPT控制的原理、过程;根据光伏发电系统的运行状况,分析并选取合适的DC/DC(BOOST)电路。然后,研究各种MPPT控制的算法和思路,尤其对控制周期和步长的选取,做出了详细的理论推导;分别在MATLAB中建立光伏MPPT控制系统模型;分析各种算法的使用场合及优劣势,并在此基础之上提出基于蛙跳粒子群神经网络的智能控制算法。之后,详细介绍了神经网络和蛙跳粒子群算法的原理和优化思路,并在MATLAB中编写控制算法,搭建了基于蛙跳粒子群神经网路的MPPT控制系统模型,仿真结果表明,此智能算法在动态响应速度、稳态精度及抗干扰能力等方面均优于传统控制方法,控制效果突出。
[Abstract]:In order to solve the pressure of environmental pollution and energy shortage, solar energy as the most environmentally friendly, the most convenient, sustainable new energy, has great potential for development. However, the high investment cost and low conversion efficiency of photovoltaic system are the main factors that affect the development of photovoltaic industry. How to improve the conversion efficiency of photovoltaic power has become the focus of attention and discussion. Therefore, as one of the important means to improve energy efficiency, photovoltaic maximum power point tracking technology emerges as the times require. Through the analysis of photovoltaic cells and various control algorithms, it is found that in the tracking control process, some traditional control algorithms are difficult to achieve an effective balance between dynamic response speed and steady-state precision, and the tracking efficiency is not ideal. In this paper, we creatively apply the leapfrog particle swarm optimization neural network algorithm to the photovoltaic maximum power point tracking control, and make use of the strong self-learning ability, nonlinear fitting ability and fault-tolerant ability of the intelligent algorithm. Combined with constant voltage control, the tracking efficiency is greatly improved. The principle and characteristics of photovoltaic cells are analyzed in detail, and a battery model is established in MATLAB to simulate the effect of changing environment on the U-IU-P characteristic curve of the cell. It provides the foundation for the modeling of photovoltaic control system. The principle and process of MPPT control are analyzed, and the appropriate DC / DC boost circuit is analyzed and selected according to the operating condition of photovoltaic power generation system. Then, the algorithms and ideas of various MPPT control are studied, especially the selection of control cycle and step size, the detailed theoretical derivation is made; the PV MPPT control system model is established in MATLAB; the use situation and advantages and disadvantages of various algorithms are analyzed. On this basis, an intelligent control algorithm based on leapfrog particle swarm optimization neural network is proposed. Then, the principle and optimization idea of neural network and leapfrog particle swarm optimization algorithm are introduced in detail, and the control algorithm is written in MATLAB, and the MPPT control system model based on breaststroke particle swarm optimization neural network is built. The simulation results show that, The intelligent algorithm is superior to the traditional control method in dynamic response speed, steady-state precision and anti-jamming ability, and the control effect is outstanding.
【学位授予单位】：中国地质大学(北京)
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM615

【相似文献】