The doctoral thesis covers the development of hybrid integration of different models for the purpose of dynamic substructuring. When dealing with large-scale products, optimal efficiency can be achieved by decomposing the whole structure to smaller substructures, applying separate analysis and later based on boundary conditions couple partial results back together. The applicability of this approach is conditioned by the ability to obtain real dynamic properties of the system for all six degrees of freedom, the definition of joints characteristics and the compliance of dynamic residues between individual models. In this thesis the implementation of the direct rotational accelerometer and in this way the expansion of the basic formulation based on translational responses to the overall response model. Additionally included rotational degrees of freedom define a physically consistent behavior of contacts, and thus extensions from current rigid point connections to flexible line contacts without any structural modifications. The main contribution relates to the development of a hybrid process in which the integration of different models improves the ability to predict dynamic responses of coupled systems. The proposed methodology introduces the combination of experimental and numerical degrees of freedom by maintaining measured dynamic properties from the experiment and expansion to the rest of unmeasured degrees od freedom. Proposed hybrid process is demonstrated in a real case study, where obtained results present improvements compared to typical dynamics substructuring methods.