Int J Performability Eng ›› 2021, Vol. 17 ›› Issue (6): 519-527.doi: 10.23940/ijpe.21.06.p4.519527

Previous Articles     Next Articles

Analysis of the Criteria for Assessing the Forecast Quality of Industrial Safety Indicators of Enterprises

Leila M. Bogdanovaa, Sergey Ya. Nagibinb,*, and Alexander S. Chemakina   

  1. aPlekhanov Russian University of Economics, Moscow, 117997, Russian Federation;
    bMoscow Aviation Institute (National Research University), Moscow, 125993, Russian Federation
  • Contact: * E-mail address:

Abstract: The paper describes changes of paradigm management in enterprise’s industrial safety based on a risk-based approach. This method includes identifying, analyzing, forecasting the risks of industrial accidents as well as assessing the possible extent of their consequences. All calculations should be made in a timely manner to take the necessary measures to prevent accidents. The technological complexity of enterprises is a complex system consisting of various components. The risks of failure of these components pose a security risk. Critical enterprises monitoring systems are introduced in order to increase the level of safety. These systems monitor the state of the nodes of the technological complex and predict the behavior of key indicators for a given period. The effectiveness of industrial safety management depends on the accuracy of these predictions. Various methods can be used to assess forecast. The purpose of this paper is to analyze the criteria used to assess the quality of forecast models and recommendations for their usage in practical life. Various coefficients based on residuals are used to assess the applicability of a particular model. As shown in the results of our research, not all indicators should be used in industrial safety systems. This paper describes the results of the possibility for using various coefficients in training forecast models and for demonstrating an empirical assessment of the model.

Key words: estimation of forecasting results, integral risk indicator, industrial safety, mathematical modeling, risk-based approach, time series analysis, time series forecasting