In the paper a new class of identification problems of analog tecnical objects, based on usage of symbolical combinatory models (SC models) is described. SC models are used as an efficient computing approach that is well suited for solving ill-conditioned equation systems, which arise in identification algorithms. On their basis, alternative methods of identification were developed that use decoding operation for obtaining estimates with clear physical interpretation. SC models are offered to use for proving analytical expressions, which allowed to discover a number of new results in the theory of identification. This way the inapplicability of stochastic models in practice of identification was demonstrated. Formalized methods for creating parallel identification algorithms were developed using SC models. That allows to increase their performance for identification of technical objects operating in normal functioning modes.