Orthogonal Test and Support Vector Machine Applied to Influence Factors of Coal and Gangue Separation

Abstract

In order to find the influence factors on coal and gangue separation, the article put forward an optimization design based on a support vector machine (SVM) and orthogonal experimental method. The crushing rates of coal and gangue were chosen as investigation target, and a comprehensive analysis was made for the coal gangue hardness relative difference value, crushing velocity, and crush frequency. According to the theory of selective crushing of coal and gangue, an impact crusher was used in crushing. The analysis showed that the relative hardness value of coal and gangue and the crushing velocity had significant effects on the coal crushing rate, while the crushing times had an insignificant influence on the coal crushing rate. The best technology scheme of coal and gangue crushing is obtained: The crushing velocity is 14 m/s, the relative hardness value of coal and gangue is 3.6, and the crushing number is 1. Relationships of various factors with crushing rate were obtained by analysis through SVM. The coal crushing rate error of experiment value and fitting value is less than 5\\percent; the gangue crushing rate error of experiment value and fitting value is less than 10\\percent. As can be seen, the SVM model has important guiding significance and high practical value.