The expert system for assessing fire risks of electrical installations in the agrarian industrial complex based on neural networks
Abstract The article is devoted to the analysis of fire risks of the operation of electrical installations in the agro-industrial complex of the region. According to statistics, the number of fires occurring for electrical reasons is steadily increasing, which makes the issue of fire risk management relevant. In order to identify and prevent these risks, a technogenic safety system has been developed, presented as a set of methods and tools, which are based on the generation of new data on the causes of threats to the operation of electrical installations based on the analysis of data on the current state of the system. Highlighted characteristics such as Human Parameters, Electrical Installation Parameters and Environmental Parameters. Mathematical models of the given components are considered, a fire risk assessment tree is compiled, in which the input parameters and intermediate vertices with the solution methods in them were determined. The developed model can be used to carry out experiments in order to study the behavior of an electrical installation under various conditions. The application of this method is especially important in cases where the removal of indicators and testing of the operation of an electrical installation is impossible due to the risk of injury or material damage.