| 摘要: | 本研究將兩組單向天鉤主動隔震系統以正交堆疊，拓展為雙向主動隔震機構。由於正交架構可假設兩向互相獨立不耦合，故進行數值模擬分析時使用單向模型以簡化過程，推導運動方程與控制律也以兩項獨立方式進行。天鉤主動隔震系統根據傳統天鉤控制理論，改良其控制力計算方式，將回饋訊號設為系統相對地表速度，前饋訊號則由地表加速度經積分濾波器求得地表速度，形成混合型前饋-回饋控制架構。此設計提升訊號取得的可行性與穩定性，並兼顧系統固有阻尼的影響。此外，控制力中加入控制力濾波器，用以濾除非預期的高頻干擾，並整合積分濾波器與控制力濾波器的影響，進行整體控制增益之最佳化設計。本研究以直接輸出回饋方式進行最佳化設計，透過參數迭代使系統絕對加速度最小化，達到最佳隔震效果。經由特徵分析發現，施加主動控制力可使系統固有頻率遠離地震主要頻率區間，降低共振風險。從頻率反應函數分析可看出天鉤主動隔震系統在地震作用頻段內優於被動系統。穩定性分析指出，系統穩定性對相對地表速度增益參數較敏感，地表速度前饋增益參數則不影響系統穩定性，且控制力延遲時間亦為穩定性關鍵因素，須納入分析。地震歷時分析之歷時，根據AC156規範於頻率域進行調整，並整理相關調整流程。本研究聚焦於一高科技廠房，調整八筆地震歷時，使其滿足地表需求反應譜及經樓層放大之需求反應譜兩種情形，作為模擬及實驗之輸入震波。為驗證數值模擬之可靠性，進一步透過簡諧震波與符合 AC156 規範之震波實驗，並比較數值模擬與實驗分析結果，結果大致呈現一致的趨勢，證實本系統與控制策略於各種外力擾動情況下具高度可行性與可靠性。本研究亦說明AC156振動台試驗準則規範，介紹其適用範圍、需求反應譜、測試反應譜與合格標準等，並討論使用AC156評估隔震系統之適用性：以AC156檢核隔震系統時，評估加速度隔震效果應為適當，若無完全包絡需求反應譜，作為隔震系統位移之評估依據則可能不足。;This study extends two uniaxial skyhook active isolation systems into a biaxial configuration through orthogonal stacking. Owing to the orthogonal arrangement, it is assumed that the two directions are independent and uncoupled, allowing the use of a uniaxial model for numerical simulations to simplify the analysis. Accordingly, the equations of motion and control laws are derived independently for each direction. Based on the traditional skyhook control theory, the control force calculation is modified by using the relative velocity between the system and the ground as the feedback signal, while the feedforward signal is obtained by integrating the ground acceleration through a low-pass filter to estimate ground velocity. This forms a hybrid feedforward-feedback control system, which improves the feasibility and stability of signal acquisition while considering the effect of inherent system damping. Additionally, a control force filter is introduced to eliminate undesired high-frequency disturbances. The combined effects of the integrator and control force filters are incorporated into the optimal design of the overall control gain. The system is optimized using a direct output feedback approach, where parameter iteration is employed to minimize the absolute acceleration of the system, thereby enhancing the isolation performance. Eigenvalue analysis reveals that applying active control forces shifts the system’s natural frequencies away from the dominant earthquake frequency range, reducing resonance risk. Frequency response function analysis further confirms that the skyhook active isolation system outperforms passive systems within the seismic frequency band. Stability analysis indicates that the system is highly sensitive to the gain of the relative ground velocity, whereas the gain of the feedforward ground velocity has little impact on stability. Moreover, the time delay in control force application is identified as a critical factor affecting system stability and must be considered in the analysis. The input ground motions for time history analysis are adjusted in the frequency domain following the AC156 requirements, with the adjustment procedure systematically documented. Focusing on a high-tech facility, eight ground motion records are modified to satisfy both ground-level and floor-amplified demand spectra, serving as inputs for numerical simulation and experimental testing. To verify the reliability of the numerical simulation, both harmonic and AC156-compliant seismic experiments are conducted, and their results are compared with the simulations. The outcomes exhibit generally consistent trends, validating the feasibility and reliability of the proposed system and control strategy under various external disturbances. Furthermore, this study explains the AC156 shake table testing criteria, including its scope of application, demand spectra, test spectra, and qualification standards. The applicability of AC156 to the evaluation of isolation systems is also discussed. While AC156 is appropriate for assessing acceleration reduction, it may be insufficient for evaluating isolation system displacement if the test spectra do not fully envelop the demand spectra. |
