Scenario–based knowledge model and language for situation assessment and prediction

Abstract

This paper is devoted to the problem of situation assessment and prediction in application where it required to have a flexible means to change the scenario behavior, depending on the progress of the system and the current state of the environment in real time. Paper reviews and analyzes the advantages and disadvantages of existing workflow languages and it is shown that the traditional language of the specification that can represent primarily reactive behavior, lack the expressive possibilities and then not be able to cope with a task. Features of the problem specification, evaluation and prediction of the situations shown in the task management system, a fragment of rocket fuel starting complex. The paper proposes a new language, which is intended to describe the knowledge of the scenarios to assess the current state of performance scenario, to predict of its development, and to select the option to continue depending on the progress of states and the state of the environment. A description of the basic elements, graphical notation and operational semantics of language are presented. Developed language capabilities are demonstrated by the example of describing a model diagnostic of unexpected situations in the process of a fragment of fuel. This application presents examples of process specifications developed in terms of language scenarios.