Decision Support Level in Monitoring Systems for Railway Automation Based on Questionnaire Theory

Abstract

The authors pay attention to the monitoring technologies development for complex technical systems. Such systems include many heterogeneous moving and stationary objects. The systems are equipped with both built-in and external means of technical diagnostics and monitoring, which form informational messages about the measurement results and the current state of the components. This information allows service personnel to promptly prevent a shutdown of the process and identify subcritical conditions. This increases the fault tolerance of the objects being diagnosed. When organizing systems for technical diagnosis and monitoring, it is often impossible to provide the necessary completeness and depth of diagnosis for making an accurate diagnosis and subsequent forecasting. However, the obtained information allows the formation of many diagnostic features inherent in the states of the objects being diagnosed. This information may be the source for the implementation at the software level of decision support subsystems by service personnel operating diagnostic objects. To form the initial data for decision support subsystems, the authors proposed to use the following data: data from measuring subsystems, historical data about the diagnostic object, statistical indicators from monitoring systems in automatic mode. As statistical indicators, the probabilities of occurrence in the various components of object defects and data on the diagnosing implementation costs are used. These data vary depending on the service life, the diagnostic object importance for the process, its effect on the system readiness, etc. The source data is used at the software level for the diagnostic algorithm’s implementation in the questionnaires form. The questionnaire is a tree-based weighted oriented graph. The output contains the recommended sequence of operations for testing the diagnostic object. That allows you to achieve the most effective localization of the defect. An example of the technologies developed for monitoring critical facilities from the railway automation field is given.