混沌改进群体智能算法研究及其在光伏MPPT中的应用

发布时间：2018-07-12 17:20

本文选题：复杂应用环境 + 光伏最大功率点跟踪　；参考：《南昌大学》2017年硕士论文

【摘要】：群体智能算法在解决具有多局部极值点的优化问题时,具有十分明显的优势,因此被广泛的使用,但由于算法本身存在一些弊端,也使得群体智能算法的应用受到了局限。本文针对群体智能算法普遍具有的“早熟”弊端,提出了一种混沌改进群体智能算法,并以人工蜂群算法和猫群算法为例,根据算法的原有数学模型提供了改进说明。仿真以光伏为背景,搭建符合实际复杂情况下的光伏系统模型,将改进后的算法应用在局部阴影情况下的光伏最大功率点跟踪,并分别进行了静态特性分析和动态特性分析。仿真结果很好地说明了所提算法能够解决原算法的“早熟”问题,具有较高的精度和更快的收敛速度,且算法的鲁棒性较好。最后通过将改进后的人工蜂群算法以及改进后的猫群算法与粒子群算法进行对比,验证了所提算法的优越性。为进一步证实算法的有效性和高效性,搭建了光伏系统小功率实验平台,编写了改进前后算法的程序,得到实验当天跟踪到的最大功率及其对应的输出电压、电流。实验结果很好地说明了改进前后的算法均能根据光照强度的改变快速跟踪到新的功率点,但本文所提算法的跟踪精度更高,获得的平均功率值更大。
[Abstract]:Swarm intelligence algorithm has obvious advantages in solving the optimization problem with multi-local extremum, so it is widely used. However, due to some shortcomings of the algorithm itself, the application of swarm intelligence algorithm is also limited. In this paper, a chaotic improved swarm intelligence algorithm is proposed to overcome the "premature" disadvantage of swarm intelligence algorithm. Taking artificial bee colony algorithm and cat swarm algorithm as examples, an improved explanation is provided according to the original mathematical model of the algorithm. The simulation takes photovoltaic as the background, sets up the photovoltaic system model in accordance with the actual complex situation, applies the improved algorithm to the photovoltaic maximum power point tracking under the local shadow, and carries on the static characteristic analysis and the dynamic characteristic analysis respectively. The simulation results show that the proposed algorithm can solve the "premature" problem of the original algorithm, has higher accuracy and faster convergence speed, and has better robustness. Finally, by comparing the improved artificial bee swarm algorithm and the improved cat swarm optimization algorithm with the particle swarm optimization algorithm, the superiority of the proposed algorithm is verified. In order to further verify the effectiveness and efficiency of the algorithm, a small power experimental platform of photovoltaic system was built, and the program of the improved algorithm was compiled. The maximum power and the corresponding output voltage and current were obtained on the day of the experiment. The experimental results show that the improved algorithm can quickly track the new power points according to the change of illumination intensity, but the tracking accuracy of the proposed algorithm is higher and the average power value is larger.
【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】