Simulation modelling, supported by System Dynamics and intensive use of modem digital computers, which implies also the mass application of, nowadays inexpensive, but at the same time very powerful personal computers, is one of the most convenient and most successful scientific methods of investigation of performance dynamics of non - linear and very complex scientific, technical and organisational systems. System dynamics uses qualitative and quantitative simulation modelling, and the basis of optimisation of dynamic systems and processes is based on the so called "heuristic" procedure, which means on the method of "manual and iterative" procedure, which has been automatized by fast digital computer, and is called "heuristic optimisation" (retry and error). The purpose of this thesis is to demonstrate the achievements of system dynamics simulation modelling while investigating the performance dynamics of a complex non - linear ship turbine generator system. The ship turbine generator system will be presented as a set of non - linear differential equations, or continuous simulation model of higher level, and it will simultaneously be a discrete simulation model. In the end various propulsion states of the system will be simulated. An efficient application of scientific methods of investigation of complex dynamic systems will be presented by: 1. qualitative method 2. quantitative simulation methodology of System Dynamics Modelling Approach - MIT and all data will be presented to experts, with the possibility of collecting additional data about the same system, in fast, scientifically based method of investigating a complex system. This means: Do not simulate performance dynamics of complex systems by investigation of the "black box", because practice in research and designing complex systems conjirmed that it far better to simulate by applying the investigation approach of the "white box ", i. e. " The System Dynamics Methodology