電離層電腦斷層掃描技術已發展多年,其主要是利用數個地面接收站接收無線訊標衛星訊號而獲得電子密度沿衛星和地面接收站射線上的線積分,即全電子含量資料,並藉此交錯射線的全電子含量資訊來反推電子密度在空間中的結構分布。基本上,這是在求解一組線性方程式的問題,然而因為電離層斷層掃描的諸多限制,使得此相關的線性方程式組,有無窮多組解甚或無解。以過學者曾使用各種電離層模型之電子密度為初始猜測值以ART 及MART 等迭代演算法來重建電子密度的結構,但需給予接近真實答案的初始猜值方能收斂到較正確的答案,也就是說解的正確與否,與使用的電離層模式息息相關。本文嘗試以另一種選擇最佳化答案的方法─二次規劃法,來求解斷層掃描的問題。先以模式模擬的資料做反演研究,再對真實資料做反演,同時並有新竹電離層探測儀的資料可供參考。我們認為以現階段多為近乎垂直射線的資料限制下,二次規劃法反演的結果因與初始猜值無關,且此方法所用之參數值,只要設定在合理範圍內,與結果相關甚少,因此改良式二次規畫法在電離層電腦斷層掃描反演法之應用上較ART 或MART 為佳。 The problem of ionospheric tomography involves finding the solution of a linear set of quations representing the integrated TEC along a large number of intersecting ray paths. Due to the limits in the ionospheric tomography, the solution to the problem is non-unique. A new independent method to select the optimal solution for reconstructing tomographic images of the ionosphere is described. The mathematical approach makes use of the quadratic programming method. Simulations and examples are presented using experimental data from radio beacon satellites for which ionosonde data are available. With the characteristic that most of the data comes from almost vertical ray paths, we consider the results from quadratic programming method are better than that fromART or MART.