EN
In this paper concept of coincidence of variable and methods for checking coincidence of model and variables are presented. Particularly Hellwig's hypothesis and methods for constructing model with difference compensators are described. It makes possible keeping non coincidentional variables in model.(original abstract)