近年来，随着各国大地震的接连发生，对人类的生命财产造成了巨大的损失，高层建筑的隔震抗震引起了广泛关注。因此，在实际结构中对与建筑隔震性能的研究具有重要的意义。本文提出依次采用扩展卡尔曼识别结构响应和最小二乘识别未知激励的方法，对隔震层的无模型非线性特性进行识别。首先是在小地震线性情况下识别出结构参数刚度和阻尼，然后在大地震下对隔震特性进行识别。算例表明,该方法对已知地震激励下的隔震结构，其非线性特性的识别具有较高的精度。这样可通过结构迟滞力的变化，对结构的隔震性能进行有效的识别。-Recently, as severe earthquake occurred one after another in different countries and caused huge losses of human life and properties, high buildings isolation and resistance have received great attention. Therefore, studies on buildings isolation performance in actual structure are significant. In this paper, an approach based on the extended Kalman estimator and least-squares estimation of unknown excitation is proposed to identify the nonlinear property of isolation layer without theoretical model. Firstly， structure parameters such as stiffness and damping are identified under minor earthquake. Then, nonlinear properties of isolation layer are identified under severe earthquake. The simulation results demonstrate that the proposed approach is capable of identifying the nonlinear property with good accuracy. So the isolation performance can be effectively recognized through the changing of the hysteresis force.