The original methodology of the system analysis of the inhomogeneous problem is offered, including stages of its reducing to homogeneous parts and selecting for them appropriate toolkits: methods and models. This system applies the accumulated knowledge and the experts skills to refer of each homogeneous problem to one or several alternative classes of modelling methods: analytical methods, statistical methods, artificial neuronets, knowledge based systems, fuzzy systems, genetic algorithms. The knowledge base testing has shown sufficiency and consistency of knowledge for realization of the inhomogeneous problems analysis even in conditions with a low and average distortion in the problem descriptions.
A Proposal of Knowledge-Based Systems Using Fuzzy Rules and Genetic Algorithms