為了將理論推展應用並進行參數研究，本文使用9座模型進行識別試驗，共有3種高度，5種斷面深寬比，實驗組合為15組，因此結果總計有15組不同之氣動力導數比較。值得一提的是，有別於林勝偉的論文，本文進行曲線擬合時同時對七個風速下之轉換函數進行擬合，解決了由個別風速擬合所產生的不同氣動力導數問題；另外，本文亦提出以梯度法配合基因演算法求解，以達到更佳的擬合結果。根據15組識別結果顯示，順風向氣動力阻尼幾皆為負值，且較高之建築具較大氣動力阻尼；氣動力勁度之正負不定，但對氣彈效應貢獻不大。 This thesis investigates the aero-elastic behavior of a series of high-rise buildings with different shapes by using a novel identification scheme that employs the forced actuating technique. The aero-elasticity of the buildings are defined by the frequency-dependent aerodynamic damping and stiffness, and they were identified through wind tunnel experiments and parametric comparisons were finally made. By following the formulation in the thesis of Mr. Lin, the relation between the motion-induced moment and the rotation angle was firstly assumed to be linear. The frequency domain representation by the parameters to be determined can then be converted into a state space equation in time domain. The incorporation of such a relation with the equation of motion under wind flow and forced actuation further leads to an aero-elastic state space equation with an input from the forced actuation. The frequency response function thus induced from this aero-elastic state space equation can be used to compare with the experimental data by curve-fitting each other in order to determine the unknown parameters and consequently the aerodynamic damping and stiffness. In performing the curve-fitting, the genetic algorithm and traditional gradient method were used in corporation to fine tune the final results. The parametric study of building aero-elasticity was conducted by using nine building models in the wind tunnel tests, which results in totally fifteen sets of results. It is worth noticing that, unlike the way employed in the thesis of Mr. Lin, this research used the experimental data under all wind speeds simultaneously in curve-fitting, thus avoided the result inconsistency from every single data set. In addition, the traditional gradient method was also proposed in this research to improve the accuracy of curve-fitted results. According to the fifteen sets of identified results, it is observed that the aerodynamic dampings in the along-wind motion are all negative, and the value increases with the building height. However, the aerodynamic stiffness could be negative or positive, and their contribution to the building aero-elasticity is not significant.
