A training computer complex has been designed for the technical and management staff of a railcar depot. Interactive use of the studied objects is implemented for the complex. A module of time series identification is integrated in the program complex. We have described a method and an algorithm for identifying time process models and time series prediction algorithms. The expression obtained gives the solution of the formed filter identification. The obtained time series models are used in the control and forecasting systems. Adaptation of the measuring and computing identification system ensures the accuracy of the forecasting assessments. Management decisions for using the material and technical resources will thus be optimal.