基于MATLAB人工神经网络预测太阳耀斑级别的研究

发布时间：2018-07-03 20:48

本文选题：太阳耀斑 + MATLB工具箱　；参考：《河南师范大学》2014年硕士论文

【摘要】：太阳耀斑是最剧烈的太阳活动。耀斑爆发时，太阳喷射出的大量的高能粒子到达地球附近时，能够影响到在地球轨道区域正常运行的人造卫星；同时国际空间站的宇航员的人身生命也面临着危险。耀斑爆发所释放出的X射线及其紫外射线，到达地球大气层中的电离层，能够破坏电磁通讯；无线电通信特别是短波通信，如手机通话、电台广播，，等会受到干扰甚至中断。采取有效的方法研究和预测太阳耀斑的级别，对于减少和避免太阳耀斑爆发给人民带来的财产和生命上损失具有很重要的意义。由于太阳活动的复杂与频繁，太阳耀斑的爆发与诸多因素有关，而这诸多因素又与太阳耀斑级别的预测有着复杂的非线性关系。采用具有较强的非线性逼近能力的人工神经网络预测太阳耀斑级别，能够综合考虑到各方面的因素，具用客观性和有效性。由此本文提出了采用BP神经网络算法进行太阳耀斑级别的预测。由于BP神经网络模型的实现需要借助于计算机编程语言，实现起来比较困难，而MATLAB软件工具箱功能强大可以解决这一问题。本文通过调用MATLAB的神经网络工具箱建立了太阳耀斑级别预测的BP神经网络模型，确定了VLF在电离层中发生SPA事件的影响因素与耀斑爆发级别之间的联系。并通过对新乡市在1998年一月到六月期间观测记录的Alpha甚低频(VLF)导航系统信号传播发生异常时的65组数据进行训练仿真，并进行预测检验。实验结果证明了该模型用于耀斑级别预测的有效性，具有很好的应用价值。
[Abstract]:Solar flares are the most intense solar activity. When the flares erupt, a large number of high-energy particles ejected by the sun can affect the normal operation of artificial satellites in the Earth orbit area when they arrive near the earth, and the life of astronauts on the International Space Station is also in danger. The emission of X-ray and ultraviolet rays from flares, reaching the ionosphere in the Earth's atmosphere, can destroy electromagnetic communications; radio communications, especially shortwave communications, such as cell phone calls, radio broadcasts, etc., can be interfered with or even interrupted. It is of great significance to study and predict the level of solar flares in order to reduce and avoid the loss of property and life caused by solar flares. Because of the complexity and frequency of solar activity, the eruption of solar flares is related to many factors, and these factors have a complex nonlinear relationship with the prediction of solar flares. The use of artificial neural networks with strong nonlinear approximation ability to predict solar flares can comprehensively take into account various factors and is objective and effective. In this paper, BP neural network algorithm is used to predict solar flares. Because the realization of BP neural network model needs to be realized by computer programming language, it is difficult to realize it, but MATLAB software toolbox can solve this problem. In this paper, the BP neural network model for predicting solar flare level is established by using the neural network toolbox of MATLAB, and the relationship between the influence factors of SPA event in the ionosphere of VLF and the flare burst level is determined. The 65 sets of data of Alpha very low frequency (Alpha) navigation system which were observed from January to June 1998 were trained and tested. The experimental results show that the model is effective in predicting flares and has good application value.
【学位授予单位】：河南师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P182.52;TP183

【参考文献】