Prediction of silicon content in hot metal based on golden sine particle swarm optimization and random forest

Abstract

Particle Swarm Optimization (PSO) algorithm quickly falls into local optimum, low precision. In this paper, add the golden sine operation to the particle position update. The results show that the improved PSO algorithm has better optimization ability. The main parameters affecting the silicon content in hot metal are selected. Then, calculate the correlation coefficient and significance level between parameters and silicon content in hot metal. Finally, the prediction model of silicon content in hot metal is established based on the Random Forest (RF) optimized by improved PSO. The results show that the hit rate is 87,17 %.

Keywords

• blast furnace
• hot metal
• silicon
• particle swarm optimization
• golden sine algorithm
• random forest