FORECASTING PROCESSES OF HIGH VOLTAGE ELECTRIC DISCHARGE SYNTHESIS OF TITANIUM CARBIDE USING MACHINE LEARNING METHODS

Abstract

The surfaces of the plasma temperature distribution in the discharge channel, the values of the pressure in the discharge channel, the pressure on the chamber wall, the amount of titanium carbide formed during processing depending on the interelectrode gap and the number of pulses are obtained by mathematical modeling using machine learning methods for spark discharge at the concentration of titanium powder in kerosene of 0.07 kg / dm³, the pulse repetition frequency of 0.3 Hz and the energy of single impact of 1 kJ. The possibility of using machine learning methods to predict the processes and results of high-voltage electric discharge treatment of titanium powder in kerosene using spark discharge and the need to consider other, more accurate machine learning algorithms, are shown.

Key words: titanium carbide, high voltage electric discharge, plasma, kerosene, machine learning, logistic regression, Random forest method.