在三相三線併網變流器之電壓支撐研究中，既有方法多著重於欠壓故障情境，對於過壓故障情境之討論相對有限。此外，多數逐相電壓支撐控制策略建立於同步單元與正、負序分量提取架構上，需依賴序分量提取器取得電壓幅值與相角資訊，使控制系統於弱電網或故障暫態下可能受到濾波延遲、參數設計與控制耦合影響。另一方面，傳統電壓判斷與限流機制常需仰賴方均根值（root-mean-square value， RMS）計算，其動態響應易受到RMS計算窗口限制，其次，故障期間之功率漣波亦影響電網之電流品質。為改善上述問題，本文提出一種基於擴充pq理論（Extended pq Theory， EPQ）之三相三線併網變流器電壓支撐控制策略。所提方法利用各相電壓峰值作為電壓支撐判斷依據，在不使用同步單元與正、負序分量提取之前提下，直接於自然框下產生逐相實功與虛功電流命令。在電壓支撐部分，本文依據電網規範與各相PCC電壓狀態產生未補償之虛功電流命令，並且為避免故障期間相電流超出額定值，本文之限流機制優先保留電壓支撐所需之虛功電流容量，再依剩餘容量決定可輸出之實功命令。除此之外，本文提出基於實功分配權重之虛功支撐補償機制，依據各相電壓故障程度重新分配實功承擔，使故障相降低實功輸出並釋放電流容量，將其容量轉換為額外虛功電流量，以提升欠壓支撐與過壓抑制效果。同時針對故障期間之二倍頻功率振盪，本文於擴充pq理論架構下設計功率漣波抑制電流命令，並加入功率漣波抑制參與權重，使正常相優先承擔功率漣波抑制能力，而故障相優先承擔電壓支撐能力。最後，本文建置一台輸出額定實功率2kW之三相六開關LCL併網變流器進行實驗驗證。其結果顯示，所提方法可於非對稱故障下實現無同步單元、無正負序提取之獨立相電壓支撐，提升故障發生時之電壓支撐動態響應，並於限流前提下兼具實功分配補償、功率二倍頻漣波抑制與電流品質改善能力，並利用變流器導納模型，說明弱電網下，所提控制之控制穩定性。;In research on voltage support for three-phase three-wire grid-connected inverters， existing methods have primarily focused on undervoltage fault conditions， whereas studies on overvoltage-fault scenarios remain relatively limited. In addition， most phase-specific voltage support strategies are established based on synchronization units and positive- and negative-sequence component extraction， requiring sequence component extractors to obtain voltage magnitude and phase-angle information. Consequently， the control system may be affected by filtering delay， parameter design， and control coupling under weak-grid conditions or during fault transients. Moreover， conventional voltage detection and current-limiting mechanisms commonly rely on root-mean-square (RMS) calculations， whose dynamic responses are inherently constrained by the RMS calculation window. Furthermore， power ripples during grid faults may also deteriorate the grid-current quality. To address the aforementioned issues， we propose a voltage support control strategy for three-phase three-wire grid-connected inverters based on the Extended pq Theory (EPQ). The proposed method employs the peak value of each phase voltage as the criterion for voltage support and generates phase-specific active- and reactive-current references in the natural reference frame without using a synchronization unit or positive-/negative-sequence component (PNSC) extraction. For voltage support， uncompensated reactive-current references are generated according to grid-code requirements and the voltage condition of each phase at the point of common coupling (PCC). To prevent the phase currents from exceeding their rated values during grid faults， the proposed current-limiting mechanism gives priority to preserving the reactive-current capacity required for voltage support， while the allowable active-power command is determined according to the remaining current capacity. In addition， this thesis proposes a reactive-power support compensation mechanism based on active-power allocation weights. The active-power contribution of each phase is redistributed according to its voltage-fault severity， such that the active-power output of the faulted phase is reduced to release additional current capacity， which is subsequently utilized to generate additional reactive current， thereby enhancing both undervoltage support and overvoltage suppression. Meanwhile， to mitigate double-line-frequency power oscillations during grid faults， a power-ripple suppression current reference is designed within the EPQ framework. A power-ripple suppression participation weighting factor is further introduced so that the healthy phases preferentially contribute to power-ripple suppression， whereas the faulted phases prioritize voltage support. Finally， a three-phase six-switch LCL-filtered grid-connected inverter with a rated active-power output of 2 kW is constructed for experimental validation. The experimental results demonstrate that the proposed method can achieve independent phase voltage support under asymmetrical grid faults without requiring a synchronization unit or PNSC extraction， thereby improving the dynamic performance of voltage support during fault transients. Under the imposed current limit， the proposed strategy simultaneously provides active-power allocation compensation， double-line-frequency power-ripple suppression， and improved grid-current quality. Furthermore， an equivalent admittance model of the inverter is developed to analyze the stability of the proposed control strategy under weak-grid conditions.
