Comparison of Biologically Inspired and Modeling Approaches for Predicting Possible Condition Change in Critical Occupational Workers

The aim of the work is to create effective methodological and technical tools for monitoring and predicting possible changes in the state of personnel in critical professions in real time. For this, modern technologies for registering the bioparameters of workers are analyzed, taking into account their physical activity. The dynamics of changes in the bioparameters of workers during a work shift, taking into account the accumulated individual data, is the basis for predicting the most probable changes in their condition. The study compares two approaches to forecasting. The first approach involves modeling based on the implementation of an iterative computational process. At the same time, a digital behavioral model is used that allows adaptation to the individual biometric data of employees. The second approach is based on a biologically inspired computational structure. To adjust the structure, the available data on changes in the bioparameters of workers during the work shift, depending on the level of work intensity, are used. Comparison of the efficiency of the two approaches was carried out on the basis of experimental data obtained during the trial operation of a sample of an automated information-measuring system for monitoring and forecasting in real production conditions at combined heat and power plant (CHPP) number 26 at Moscow. The second approach showed a higher prediction accuracy.