基於卡爾曼濾波建立即時性區域電離層模型－－以臺中市為例

Using Kalman Filter Algorithm To Build Real-Time Regional Ionospheric Model -A Case Study of Taichung City

本文以台中市作為研究區域，建立一個即時性的區域電離層模型以克服IGS全球電離層模型於時間及空間上解析力的不足。首先利用相位水準技術演算法提取出精度較高的電離層傾斜電子含量，再透過卡爾曼濾器波估計出電離層垂直電子含量及衛星與接收儀之硬體延遲誤差，精確量化電離層的延遲誤差。實驗第一部分比較2014年四個季節電離層變化情形，實驗結果顯示區域（台中）電離層模型與IGS全球電離層模型的變化趨勢大致相同，電離層電子含量的日平均差值約3~10TECU。第二部分將區域（台中）電離層模型與IGS全球電離層及Klobuchar模型進行單點定位測試，分析比較三種電離層修正模型定位精度，其結果顯示在電離層活動較穩定時三種模型相對修正比例彼此之間相差約5%，當電離層活動較劇烈時，IGS全球電離層與Klobuchar模型無法有效修正電離層誤差，而台中區域電離層模型仍然能保持約76%，並不會受到電離層變化影響，皆能保持一致的修正比例，也證明了利用卡爾曼濾波所建立之區域電離層較全球電離層模型更能符合區域性的使用。

This article selected the Taichung City as research area, build a real-time regional ionospheric model to overcome the global ionospheric model doesn’t have high resolving capability both in time and space. First used phase leveling algorithm to estimate more high precision STEC (Slant Total Electron Content) value. Second used the Kalman filtering algorithm to estimate ionosphere mode coefficients of each epoch and the differential code biases of receiver and satellites, through this way can quantize ionospheric delay more precisely. The first part of the experiment, compared 2014 each seasons ionospheric activity, according the result regional (Taichung) ionosphere model and global ionosphere model has approximately trend, average TEC of day difference about 3 to 10 TECU. Last perform the single point positioning test with the regional ionospheric model, global ionospheric model and Klobuchar model to compare and analyze each ionospheric model’s performance. The result of the test, during the ionosphere quiet, the positioning accuracy of each model have 5% difference with each other. While the ionosphere is extremely active, the regional ionospheric model have higher positioning accuracy than the others. Positioning accuracy can keep in 76%. It proved used Kalman filter to build up a real-time ionospheric model is an effective method and the regional ionospheric model is more suitable than global ionospheric mode.