The important function of the automatic process control system is the ability to forecast and control the quality of a product. The implementation of these functions can be done by using formal models that allow to do the quick reconfiguration of the technological process when the composition of the raw materials and the requirements of product characteristics are changed, etc. in relation to the prediction and management of quality indicators of road and construction bitumens, methods of getting formal models that relates bitumen quality indicators with technological parameters of their production both for classical regression equations and for formal neural networks. The method of getting such models has been worked out on abstract dependencies with further adaptation of the production of bitumen. According to the received formal models the automated process control system can solve both the direct problem of predicting quality indicators by substituting the input technological parameters for the model input, and the inverse problem using nonlinear programming methods to minimize the deviation functions of the quality parameters obtained from the model from the target values given operator technologist.

Keywords: Bitumen, quality indicators, quality forecast, management of the quality, formal models