Industrial Automatic Control Systems and Controllers Annotation << Back Process Control System for Separation in the Cement Production Shop Е.А. Muravieva, P.A. Sabanov In the article a system to control cement separation process using neural networks was developed. The controlled object is a separator. A task to improve the control system of cement production process through the installation of a frequency converter and an air-supply valve with an electric drive was set. In order for the current frequency and the opening rate of the valve to be set automatically, a neural network was developed which has two input and two output parameters. The neural network is able to control the separator rotor speed depending on sensors which control the rotation speed of separator rotor and the cement particle size. If the rotation speed of the separator rotor becomes low, artificial intelligence will adjust the fineness of grinding, changing the opening rate of the valve to supply air to the separator. If the rotation speed of the separator rotor approaches its maximum and the grinding fineness has not yet reached the specified values, then the grinding fineness is adjusted by changing the opening rate of the air supply valve to the separator. Keywords: Separator, grinding fineness, valve opening rate, frequency converter current frequency, control system, neural network. DOI: 10.25791/asu.8.2022.1378 Pp. 03-11.