本研究針對多自由?撓性伺服機械系統之關節??進?鑑別。本文延續原有敏感?最佳法之鑑別工作所推導的?態空間方程、測試?據，?用遺傳演算法，依據目標函?與設計變?設定範圍，進?設計變?之鑑別。並針對演算程序及結果，與敏感?最佳法之鑑別比較。文中以?個?子進???鑑別，範?之設計變?分別為6個阻尼??以及6個阻尼加5個勁???。對系統反應?考點敏感?較大之設計變?，?種演算法計算結果差?較小。

The objective of this research is to develop an identification method for the joint parameters in flexible servomechanisms. The derivation of dynamic and control equations, tests of frequency responses, and identification of joint parameters by a sensitivity optimization method have been conducted before. With the same test data, the parameters identification by means of genetic algorithms, according to the selected objective functions and given parameter ranges, is performed in this paper. Two examples are shown, one with six damping coefficients to be identified, while the other has five stiffness and six damping coefficients to be identified. The results are compared with those of the sensitivity optimization method. The comparison shows that the parameters with larger sensitivity have less discrepancy.