太阳黑子活动周期特征的神经网络和小波分析

BP Neural Network and Wavelet Analysis of Period of Sunspot Activity

太阳黑子数是描述太阳活动水平的主要指标,太阳活动直接影响日地环境。依据前人对太阳黑子数的观测资料,采用BP神经网络及小波分析和自相关相结合的方法,分析了1770-1869年的太阳黑子数年均值,得出了太阳黑子存在11-12年周期的结论,并对该算法及噪声鲁棒性进行了仿真。实验结果表明,该算法对研究太阳活动的本质规律是有效的。两种方法与其他方法,如自相关法、功率谱法等,进行了相比,不仅得出与实际一致的结论,而且对噪声有较强的鲁棒性,这对含噪信号的分析研究是很有意义的。

The sunspot number is the main indicator of the level of solar activity, solar activity directly affects the daily environment. Based on the sunspot number observation data of the predecessor, using BP neural network and wavelet analysis and self integrating method, the 1770-1869 sunspot number mean is analyzed, it is concluded that the sunspots are 11-12 year cycle, and the algorithm and its noise robustness is simulated. The experimental results show that the algorithm is effective for the essential rule of solar activity. Two methods with other methods, such as self correlation method, the power spectrum method, are compared to not only draw the practical conclusions but also have the strong robustness for noise, which is very significant for noise signal analysis.