Robustly identifying nonlinear mechanical systems is generally a challenging task, and this is particularly true when the structure under test exhibits nonlinear behaviors related to both stiffness and damping. In this work, a hybrid nonlinear identification approach is proposed by combining the restoring force surface (RFS) method and the nonlinear subspace identification method. The multiparameter nonlinear identification strategy is based on a first characterization conducted using the RFS method, followed by a nonlinear state-space representation using subspace algorithms. Two common friction simulation examples and one complex multi-degree-of-freedom system are employed to verify the proposed method. The effect of the measurement noise on the parameter estimation results is investigated by corrupting the previously generated output, adding different levels of Gaussian zero-mean noise. Results show that the nonlinear coefficients associated with the stiffness and damping nonlinearities can be identified with a high level of confidence, and the proposed method works well under different noise-level contaminations.

稳健地识别非线性机械系统通常是一项具有挑战性的任务，当被测结构表现出与刚度和阻尼相关的非线性行为时尤其如此。在这项工作中，结合恢复力面（RFS）方法和非线性子空间识别方法，提出了一种混合非线性识别方法。多参数非线性识别策略基于使用 RFS 方法进行的第一个表征，然后是使用子空间算法的非线性状态空间表示。两个常见的摩擦仿真实例和一个复杂的多自由度系统被用来验证所提出的方法。通过破坏先前生成的输出来研究测量噪声对参数估计结果的影响，添加不同级别的高斯零均值噪声。结果表明，与刚度和阻尼非线性相关的非线性系数可以以高置信度识别，并且所提出的方法在不同噪声级污染下都能很好地工作。