基于SSA-Elman神经网络的电离层TEC短期预报模型

针对基于神经网络的电离层TEC短期预报存在精度较低、易陷入局部最优的问题,利用CODE中心提供的TEC数据及地磁活动指数,建立基于麻雀搜索算法(SSA)改进Elman神经网络的电离层TEC短期预报模型,并通过BP模型、Elman模型及SSA-Elman组合模型分别对电离层平静期和扰动期中低纬度TEC进行5 d连续预报....

Short-Term Prediction Model of Ionospheric TEC Based on SSA-Elman Neural Network

We focus on the problems of low accuracy and easy to fall into local optimum in the short-term prediction of ionospheric TEC based on neural network. We use the TEC data and geomagnetic activity index provided by the CODE center to establish an improved Elman neural network model based on the sparrow search algorithm(SSA). The BP model, Elman model and SSA-Elman combined model are used to predict 5 days continuous TEC in the middle and low latitudes during the ionospheric quiet period and disturbance period. The experimental results show that when the optimized Elman neural network model is used to predict 5 days continuous TEC, the root mean square error of single day can reach 1.443 TECu, and the correlation coefficient can reach 0.976, which is better than BP model and Elman neural network model.