多星系GNSS地表變形監測網最佳化設計

Abstract

全球衛星導航系統（Global Navigation Satellite System，GNSS）已被廣泛應用於地表監測任務上，而網形布設與規劃為監測任務最基礎的工作，然而在多星系發展下，現有變形監測網設計方法並未完整考慮多星系之效益，大多仍採單一星系進行網形設計且缺乏系統化及全面性的優化策略。因此本研究針對多星系地表變形監測網進行最佳化設計，考量監測站真實地形遮蔽效應之分析方法，結合最小可偵測應變量指標做為網形優化設計依據，運用粒子群演算法建置自動最佳化變形監測網。本研究以南迴鐵路監測網資料進行實驗測試，成果顯示應用多星系GNSS衛星資料可使監測點位精度及網形應變偵測靈敏度提升至少30%。此外，藉由粒子群演算法自動最適點位搜尋，可改善網形幾何並使網形整體應變偵測能力增加比率達92.5%。透過本研究建構之變形監測網最佳化策略，能有效促進多星系GNSS衛星於地表變形監測上具更高之應用效益。

Global navigation satellite system (GNSS) has been used worldwide for ground deformation monitoring tasks. With the rapid advances in GNSS, the benefits of integrated multi-constellation GNSS are not considered in the existing network design. Additionally, the traditional method applied in optimization of network design are inefficient. In this study, the multi-constellation GNSS were applied and terrestrial obstruction effects were considered to evaluate the quality of a network positioning result. Then, the minimum detectable principal strain index was introduced to be the optimal design criteria. Finally, the particle swarm optimization (PSO) method was adopted for an automatic design of deformation monitoring network. The results revealed 30% improvement in the accuracy of positioning and deformation analysis when multi-constellation GNSS systems were applied. Moreover, an automated optimization method of the multi-constellation GNSS monitoring network configuration was achieved by PSO method. Through the optimization of the monitoring network, the minimum detectable principal strain index was improved up to 92.5%. Consequently, the profit of introducing the multi-constellation GNSS techniques in the ground deformation monitoring tasks can be maximized with the proposed approach.