中文摘要：

针对受力状态映射法作为辨识非线性结合部参数重要方法具有物理意义明确、辨识准确等特点,而实际辨识试验时结构某些参数难以测定、致辨识过程困难等问题,提出结合子结构模态综合法的受力状态映射法,可有效解决该问题。参数辨识过程中对模型易出现的缺失项及冗余项,利用基于TSVD（Truncated Singular Value Decomposition）正则化方法的模型迭代修正法。仿真模型表明,修正模型较精确模型误差小、辨识精度高。通过对导轨结合部进行参数辨识实验研究,证明辨识方法的有效性及准确性。

英文摘要：

Force-state mapping method is an important method for the parameter identification of nonlinear joints. The method is of attractive features of clear physical meaning and high identification accuracy. However, certain parameters are difficult to measure during actual identification experiments. Aiming at such problem, a force-state mapping method based on substructure modal synthesis method was proposed. Furthermore, an iterative correction method based on the truncated singular value decomposition （TSVD） regularization was proposed to overcome difficulties associated with term-deficiency and term-redundancy, which may occur during the parameter identification process. The simulation results show the high identification accuracy, demonstrated by the small error between corrected and precise models. Finally, the parameter identification of the joints of a rolling guide was carried out experimentally. The experiments verify the effectiveness and correctness of the identification method.