Abstract
The article presents a system of artificial intelligence, made for expert assessment of electrical safety level at power generation facilities. This information system, based on fuzzy logic methods, can more precisely calculate electrical resistivity of the soil in specific conditions. The researchers describe the technique of forming a database and a knowledge base of the expert system including a set of models and algorithms. The article provides a detailed description of fuzzy models of electrical resistivity $\rho$ for different types of soil depending on its humidity, salinity and temperature (both in positive and negative ranges). It offers fuzzy climatic models of soil in a specific area, taking into account changes of soil $\rho$ deep into the ground throughout a year. The created artificial intelligence system evaluates the security level of electric power facilities on the basis of different algorithms, used to calculate grounding, touch and step voltages; it can more precisely determine soil $\rho$ deep into the ground during a year. The researchers add an option of graphic depiction of modeling results for the validation of the solutions, offered by the expert system.