The article presents the investigation results of the models and the algorithms for the formation of a wide range of transport and logistics real time processes, which create sorted structures of the mass orders. Operators of the different complexity, «weight» are used in this process. Different issues related to the creation of the formal models of the input data sets are resolved. They provide an effective implementation of the technological and the logistic processes. The purpose of the models is to improve the procedures for optimal ordering and classification of the sequences of analysing elements and orders. We have proposed new specialized models (graph models, binary trees) for the input (primary) sets of the elements, as well as algorithms for their processing, which ensure an efficiency increase of the ordering process components. In addition, graph models and algorithms allow solving classification tasks for the data of various types, and they are also suitable for organizing multi-sequencial orders. The high computational efficiency of the proposed new algorithms for arranging and classifying data has been established using comparative analysis. The article provides meaningful examples and notes the peculiarities of the tasks used for real time ordering and classification of the multi-sequencial orders. Namely, this is the task of disassembling and forming railway trains and the task of «mass order delivery to address». Examples of real time creation and transformation of the data flows binary graph models are provided to demonstrate the models and the algorithms. The formed models have been also applied to the tasks of effective sorting and classification with interval uncertainty of the data. We have investigated the possibilities of fuzzy arrangement structure creation and classification of numerical data received online.

The researches results of discrete optimal planning problems of a wide range of production-technological, logistic and other service processes are presented. The planning methods are based on new intelligent procedures for ordering (IPO) sequences of elements (orders), which are implemented by means of constructive modeling. Purpose of procedures is to increase the efficiency of ordering receiving of orders, taking into account the complexity of the formation operations, as well as resource constraints. The article considers the models and methods of IPO application, which are focused on the processes of disbandment-formation (DF) of multigroup railway trains at sorting stations. Formally, such processes are represented by new models of ordering multi-sequences of orders taking into account the complexity of operations (OMSCE). In the search for optimal solutions, models of Hamming's associative memory are used, which allow to classify the current situations of OMSCE processes. In them, each class of certain states (taking into account the incompleteness and data perturbation) corresponds to one or more rational operators from among the possible ones. IPO procedures reduce the number of analysis options and increase the numerical efficiency of the optimizing multi-sequence orders method. The article presents the formalization of multilayer constructive models of OMSCE processes, intelligent procedures for methods of their implementation, the formation of the procedure for operations classification based on models of Hemming neural networks. At the same time, an improved structure of DF information technology with the use of intelligent procedures is also developed, examples of their application are given.