基于小波神经网络MPPT算法研究及应用

发布时间：2018-03-16 11:32

本文选题：太阳能光伏发电　切入点：最大功率点　出处：《东华理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着全球经济的迅速发展,环境污染与能源危机问题凸显,发展新能源能显的尤为重要。在各种新能源中太阳能以其独特的优势得到迅速的发展。在太阳能的各种运用中光伏发电备受青睐,但光伏发电普遍存在太阳能转换效率低的问题。本文就如何提高太阳能转换效率问题进行了以下研究:采用最大功率点跟踪是提高太阳能转换效率的有效途径。本文详细介绍了最大功率点跟踪技术的国内外研究现状及未来发展趋势。研究了光伏电池的等效电路并给出了光伏电池的数学模型,在此基础上分析了光伏电池的输出特性曲线。并给出了几种常用的MPPT(Maximum Power Point Tracking)是最大功率点的简称。MPPT算法包括定电压跟踪法、扰动观察法、电导增量法、模糊逻辑控制法、神经网络算法等等。在深入研究了传统算法的优缺点的基础上提出了小波神经网络MPPT算法,通过MALTAB进行实验数据处理分析,验证算法的有效性。在提出新算法的基础上对MPPT算法控制电路进行了研究,对控制电路中的DC/DC变换电路进行选型和研究,选择以BUCK电路作为DC/DC变换电路,并研究了BUCK电路中电感材料、骨架结构、绕线方式、线径等对系统能量转换效率的影响。根据MPPT控制电路的实际情况设计了PWM驱动电路来抬高PWM输出波形的基准电压,使得控制电路能够正常工作,最后利用设计的MPPT控制电路及其实验设备搭建了实验系统进行小波神经网络MPPT算法的样本数据获取。本文通过分析传统MPPT算法的优缺点提出了小波神经网络MPPT算法并结合MPPT控制电路获得测量数据对算法进行了样本训练,训练结果表明小波神经网络MPPT算法可以准确预测系统的最大功率点,能够提高太阳能的转换效率。
[Abstract]:With the rapid development of the global economy, the problems of environmental pollution and energy crisis are highlighted. The development of new energy is particularly important. Solar energy has been developed rapidly in all kinds of new energy sources with its unique advantages. Photovoltaic power generation is very popular in various applications of solar energy. However, the problem of low efficiency of solar energy conversion exists in photovoltaic power generation. In this paper, how to improve the efficiency of solar energy conversion is studied as follows: the use of maximum power point tracking is an effective way to improve the efficiency of solar energy conversion. In this paper, the research status and future development trend of maximum power point tracking technology at home and abroad are introduced in detail. The equivalent circuit of photovoltaic cell is studied and the mathematical model of photovoltaic cell is given. On this basis, the output characteristic curves of photovoltaic cells are analyzed, and several commonly used MPPT(Maximum Power Point tracking algorithms, which are referred to as maximum power points, include constant voltage tracking method, perturbation observation method, conductance increment method, fuzzy logic control method, etc. On the basis of studying the advantages and disadvantages of the traditional algorithm, the wavelet neural network MPPT algorithm is put forward, and the experimental data is processed and analyzed by MALTAB. On the basis of the new algorithm, the control circuit of MPPT algorithm is studied, the DC/DC transform circuit in the control circuit is selected and studied, and the BUCK circuit is chosen as the DC/DC transform circuit. The influence of inductance material, skeleton structure, winding mode and wire diameter on the energy conversion efficiency of BUCK circuit is studied. According to the actual situation of MPPT control circuit, the PWM drive circuit is designed to elevate the reference voltage of PWM output waveform. Make the control circuit work properly, Finally, using the designed MPPT control circuit and its experimental equipment, the experimental system is built to obtain the sample data of the wavelet neural network MPPT algorithm. This paper presents the wavelet neural network MPPT calculation by analyzing the advantages and disadvantages of the traditional MPPT algorithm. The method is combined with MPPT control circuit to obtain the measurement data, and the algorithm is trained by sample. The training results show that the wavelet neural network MPPT algorithm can accurately predict the maximum power point of the system and improve the conversion efficiency of solar energy.
【学位授予单位】：东华理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM914.4;TP183

【参考文献】