The development of information and communication technologies creates increasingly complex set of circumstances that influence research, development and innovation in various industries. Complex system research and analysis have become significant, because it is an essential condition to understand, manage and improve these kinds of systems. However, the nature of complex systems and available information are the criteria, which limit the choice of appropriate modelling approach and tool. Structural modelling approach allows to perform complex system structure modelling, but until now it was not implemented in a suitable computer system. The thesis is devoted to the improvement of structural modelling approach and development of an intelligent system for complex system structure modelling. Complex and intelligent systems, frames have been analysed and structural modelling approach reviewed to develop architecture of the intelligent system. Structural modelling syntax and semantics have been improved and knowledge acquisition and representation schema frame set developed. The schema have been implemented in the intelligent system I4S, that allows the acquisition and representation of expert knowledge and to apply it for structural modelling purposes. The developed system I4S have been practically verified performing structural modelling and analysis for complex technical system robot AGR8. The thesis consists of introduction, four chapters and conclusions. The main part of the dissertation contains 168 pages, 102 figures and 153 bibliographical sources, additionally the thesis contains 3 appendices.