A method to predict the monthly mean values of f0F2 is presented inthis paper. The analysis results of monthly values of f0 F2 indicatethat the characteristic of this ionospheric parameter change withdifferent month and different year. Based on it, the f0F2 monthlymean value is predicted by taking sufficient data of many yeas intoaccount and improved the predicting method. Compared with the conserveddata, the average error is less than 0.34 MHz. Then, the fractioncharacteristic of ionosphere has been set forth by using the theory offraction and the fraction dimension of the f0F2 monthly mean valuehas been acquired. Based on it, 3 parameters are selected to predict the f0F2 monthly meanvalue for different year, thus, the predicted technology is improvedfurther. Compared with the conserved data, the average error is lessthan 0.3 MHz.%提出了利用人工神经网络技术预测电离层临界频率月中值的方案.在利用人工神经网络技术研究电离层月中值隔月变化规律的基础上,考虑足够的周年和黑子周期变化的数据训练网络,使f0F2月中值预测值与实测数据比较平均误差为0.34 MHz,预测精度有了较大改进.最后采用分形学的基本理论得到电离层f0F2月中值的分数维为3,选用3个输入量,分别预测高、低年f0F2月中值,与实测数据比较平均误差为0.3 MHz.
展开▼