贴近度在地磁分类预报中的应用

Application of close degree to classified prediction of geomagnetic disturbances

本文运用模糊数学中贴近度、从属函数和综合评判法，结合日冕物质抛射的太阳观测和其引起的行星际太阳风暴的行星际闪烁观测，提出预测经IPS认证的行星际激波引起的地磁扰动强弱程度的贴近度法，以1997-2003年期间74个IPS认证的CME-激波引起的磁扰事件的分析研究为基础，给出了用来评判磁扰研究的5个聚类指标及它们的加权集，利用综合评判法给出了小、中、大三类事件的典型集．针对每个事件建立模糊集，根据此模糊集与典型集贴近度的大小对磁扰事件进行分类预报．初步研究结果表明，对于所研究的74个磁扰事件而言，小磁扰事件预报准确率为53．3％；中磁扰事件预报准确率为100％，大磁扰事件预报准确率为90．20％，此方法平均准确率达到84％，这是模糊数学的贴近度在地磁扰动研究中的一种有效的应用．

Using the close degree and membership function and comprehensive evaluation in fuzzy mathematics, in conjunction with the solar observation of CME and IPS observation for the interplanetary solar storms caused by CME, we present a close degree prediction method for geomagnetic disturbance events caused by coronal mass ejection （CME）-shock waves identified by IPS observation. Based on the analysis of 74 geomagnetic disturbance events caused by CME- shock waves identified by IPS observation during 1997 - 2003, five clustering indexes and their weighted fuzzy sets are constructed. Typical fuzzy sets of small, moderate, and intense geomagnetic disturbance events are built based on comprehensive evaluation. By constructing the respective fuzzy set of each event, based on the close degree between the fuzzy set and typical fuzzy sets, the prediction for different geomagnetic disturbance caused by the events can be made. Main results are as follows： For the 74 geomagnetic disturbance events, the prediction accuracy is 53.3% for small geomagnetic disturbance; 100% for moderate geomagnetic disturbance; and 90.20% for intense geomagnetic disturbance. The average precision of this method is 84%. This method is an effective application of fuzzy mathematics to studying the geomagnetic disturbance.