节点文献
支持向量机回归算法预测局部遮阴光伏发电系统最大功率
Research on MPP of Partially Shaded PV System Based on SVR Algorithm
【作者】 赵丹;
【导师】 程泽;
【作者基本信息】 天津大学 , 控制理论与控制工程, 2010, 硕士
【摘要】 太阳能是一种新型的绿色可再生能源,光伏发电产业在未来有着良好的发展前景。但是由于光伏电池能量转化率低以及伏安特性的高度非线性等特点,因此研究具有全局搜索能力的MPPT控制算法,实现光伏发电系统的最大输出功率,是提高光伏电池利用率的关键。实际的光伏系统受外界环境影响较大,系统输出存在随机性,特别是局部遮阴或光照不均等情况造成的阵列失配以及热斑现象不仅会影响光伏系统的功率输出,更造成安全和可靠性问题,并且在局部阴影条件下光伏阵列的P-V特性出现多个极值点,使得常规的最大功率跟踪算法失效。因此本文将支持向量机回归算法应用于光伏系统研究,利用支持向量机的全局优化、适应性强、泛化性能好等优点,对光伏组件的最大功率点进行预测跟踪。论文分析了光伏电池的电路模型和输出特性,在此基础之上对光伏阵列特性进行仿真,对目前国内外局部遮阴下最大功率算法的研究成果进行分析与评述。在分析了最大功率点影响因素的基础上,对支持向量机回归算法理论进行分析与阐述,针对光伏阵列特性出现的多极值问题,提出了基于支持向量机回归拟合的最大功率预测模型。仿真结果表明,支持向量机预测模型具有较高的精度和效率。在Matlab/Simulink中建立基于SVR预测模型的光伏系统最大功率点跟踪控制系统,仿真结果验证了算法的有效性。最后,搭建了系统的硬件电路,在室外全光照条件下和局部遮阴情况下进行了实验。实验结果表明系统运行稳定,对于环境的变化有较好的响应能力。
【Abstract】 Solar energy is a new type of green and renewable energy, photovoltaic power generation industry has good future prospects. However, due to low solar cell energy conversion and voltage characteristics of highly nonlinear, so the research has the global search ability of MPPT control algorithm to achieve the maximum output of photovoltaic power generation system is to increase the efficiency of photovoltaic cells key.Actual PV system influenced by the external environment, there is randomness of the system output, especially local conditions such as partially shaded or uneven illumination, these phenomenon of loss will not only affect the power output of PV systems but also leads to security and reliability problem. Further more,under partially shaded conditions, the P-V curve of PV arrays will have the characteristics of multi-summit, which makes the conventional MPPT algorithm is disabled. Support Vector Machine has the advantages of global solutions, good adaptability and high generalization ability in theory. In this paper, support vector regression algorithm is used to predict the maximum power point tracking.This paper analyzes the circuit model and the photovoltaic output characteristics,in this based on the simulation of the solar arrays,analyzed and reviewed at home and abroad under the conditions of partially shaded maximum power algorithm research.After analyzing factor affecting the change of the position of MPP,the theory of SVM regression is presented.Characteristics of photovoltaic arrays for the emergence of multi-extremal problem is proposed based on support vector machine regression prediction model for maximum power.Simulation results show that SVM model has high accuracy and efficiency.In Matlab/Simulink to establish SVR prediction model based on MPPT of photovoltaic control system,simulation results demonstrate the effectiveness of the algorithm. Finally, the hardware circuit of the control system is constructed through the use of AVR micro-processor. All experiments in outdoor lighting conditions and partially shaded under the circumstances.Experimental results show that the system is stable ,changes in the environment a better response capability.
【Key words】 Support Vector Machine Regression; Partially Shaded; Maximum Power Point Tracking; PV System;