基于模糊神经网络的光伏蓄电池剩余容量预测优化研究

发布时间：2018-04-26 18:33

本文选题：模糊神经网络 + 剩余容量　；参考：《湖南科技大学》2015年硕士论文

【摘要】：太阳能光伏系统具有绿色无污染、安全便捷、不受地域限制和储量无限等优点。铅酸蓄电池是太阳能光伏系统中的关键储能装置,因此对蓄电池剩余容量高精度预测的研究是非常有必要的。实现剩余容量的精准预测不仅有利于提高蓄电池的工作效率,有效延长工作寿命,并且能恰当的防止其过充过放。但是铅酸蓄电池内部复杂的电化学特性导致剩余容量的预测问题再业界一直是一项较难攻克的问题,对于容量预测方面也没有统一的研究标准。本文通过翻阅大量相关文献和学术期刊,在学校实验室已有的研究基础上,在现有的模糊神经网络基础上进行改进,将其运用于太阳能光伏照明系统的剩余容量预测。本论文首先叙述了铅酸蓄电池的主要特性参数等理论知识,主要包括有电池电压、蓄电池容量、荷电状态、蓄电池内阻。其次详细分析了铅酸蓄电池的放电特性和影响铅酸蓄电池寿命的四大因素,并讨论了蓄电池的工作原理及等效电路模型。接着对目前国内外铅酸蓄电池剩余容量的预测方法进行了详细总结,并比较分析各类方法的优缺点。然后考虑到太阳能光伏照明系统中铅酸蓄电池的特性,选择以模糊神经网络算法为基础的改进模型进行蓄电池剩余容量预测。设计输入层、隐含层、推理层、求和层以及输出层。其主要特点是在于对输入数据选用7个语言变量来定义模糊集合,隶属度函数选用高斯函数,推理层采用两个模糊规则分别为乘积和求和。并提出对自学习算法的改进,通过与另外三种常用算法进行实验对比分析,可知改进算法针对于改进的模型在预测精度和匹配度上起到优化作用。最后通过在MATLAB软件中进行仿真实验,从仿真数据和实验数据对比的曲线观察可知,改进的模糊神经网络模型和算法能在不影响电池正常工作的情况下对光伏蓄电池剩余容量预测精度和稳定性上有所提高,具有一定的应用价值。
[Abstract]:Solar photovoltaic system has the advantages of green pollution, safe and convenient, free from geographical restrictions and unlimited reserves. Lead acid battery is the key energy storage device in solar photovoltaic system, so it is necessary to study the high precision prediction of battery residual capacity. The accurate prediction of residual capacity can not only improve the working efficiency and prolong the working life of the battery, but also prevent its overcharging and overloading properly. However, the complex electrochemical characteristics of lead-acid batteries lead to the prediction of residual capacity, which is always a difficult problem in the industry, and there is no uniform research standard for capacity prediction. In this paper, a large number of related literatures and academic journals are reviewed, and based on the existing research in the school laboratory, the fuzzy neural network is improved and applied to the prediction of the residual capacity of solar photovoltaic lighting system. In this paper, the main characteristic parameters of lead-acid battery are described, including battery voltage, battery capacity, charge state, battery internal resistance and so on. Secondly, the discharge characteristics of lead-acid battery and the four factors affecting the life of lead-acid battery are analyzed in detail, and the working principle and equivalent circuit model of the battery are discussed. Then, the prediction methods of residual capacity of lead-acid batteries at home and abroad are summarized in detail, and the advantages and disadvantages of various methods are compared and analyzed. Then, considering the characteristics of lead-acid battery in solar photovoltaic lighting system, an improved model based on fuzzy neural network algorithm is selected to predict the battery residual capacity. Design input layer, hidden layer, inference layer, summation layer and output layer. The main features are that the fuzzy set is defined by seven language variables for input data, Gao Si function is used for membership function, and two fuzzy rules are used in inference layer for product and summation respectively. The improvement of self-learning algorithm is put forward. By comparing and analyzing with other three common algorithms, we can see that the improved algorithm plays an optimization role in prediction accuracy and matching degree for the improved model. Finally, through the simulation experiment in the MATLAB software, from the curve observation of the contrast between the simulation data and the experimental data, we can know, The improved fuzzy neural network model and algorithm can improve the prediction accuracy and stability of the residual capacity of photovoltaic battery without affecting the normal operation of the battery.
【学位授予单位】：湖南科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TM912

【参考文献】