在動態與快速靜態的衛星測量應用中，利用載波觀測量求解位置時，如何快速得到正確的整數相位模稜值，是求解精度與效率的關鍵。但係數矩陣會因為參數間高相關而須加長觀測時間，倚賴衛星幾何改變以改善。因此白化濾波(Whitening Filter)技術即將原在參數域高相關的問題，投影至另一低相關域。使數學變換過程等效於衛星幾何改變，進而可在較短觀測時段裡求解。 白化濾波是利用Crout因子分解，使一正定對稱矩陣分解為對角線矩陣與單位上下三角矩陣之連乘。應用其矩陣對角線化的特性於相位模稜實數解的協變方矩陣上，大幅減少整數相位模稜的候選解。第二部分利用二次差所求得的位置參數與整數相位模稜，估計一次差的相位模稜，並代入一次差的處理模式以求解位置與時間參數。第三部分為動態測量，利用混合平差的特性，將時變參數向量移項至殘差向量中，目的為階段性的減少未知參數數目，使動態測量在求解過程中更有效率。 In the practice of kinematic and rapid static surveying, the key point to reach the target of precision and efficiency while using carrier phase for location is how to obtain quickly accurate integers of ambiguity. However design matrix can vary because of the high correlation between parameters and longer observation time depending geometric satellite's improvement. Therefore whitening filter is a technique reflecting the high correlation in the space of parameters on another little lower correlation, causing the effects of mathematics chance turn to be the same as geometric satellite change, and get the result within a short observation period. Whitening filter uses crout factorization to decompose an positive-definite symmetrical matrix into the continue multiplication of diagonal matrix and unit upper/lower triangular matrix. Applying the specifics of its diagonal matrix condition to covariance matrix can reduce the number of solution ways for integral ambiguity. The second part uses double differences to get location parameters and integer ambiguity. Then estimates single differences ambiguity for adopting the latter process to find out parameters of location and time. The third part deals with kinematic surveying. Using characteristics of mixed model to convert vector of variance into vector of residuals with the purpose of reducing unknown parameters, so that kinematic surveying can be more efficient in the satellite surveying.
