ИСТИНА

cing is a dangerous phenomenon in aviation, transport and energy. From the point of view of a physical process, this is a complex phenomenon that includes gas dynamics, dynamics of particles, liquid film dynamics, thermodynamic and phase transitions. The shape of the ice depends on various conditions: wind velocity, the temperature of the flow, size of a droplet in air, quantity of droplets, and the time of influence of the flow on the airfoil. Flight experiments for researching icing are expensive and depend on weather conditions. Laboratory experiments require special facilities and climate wind tunnels. Numerical simulation of this process requires special software and supercomputer resources. The aim of this study is to develop a special neural network that can predict ice shape on an airfoil.