ABSTRACT

Due to the adjustment of the Chinese industrial structure in recent years, the coal industries are facing a sluggish market environment. Therefore, it is important to study the product specification prediction and economic benefit maximization prediction of coal preparation plants. First, the washability data and distribution curve models were used to calculate the quantity and quality index of the products of the shallow slot heavy medium separator, the heavy medium cyclone, and the spiral separator, respectively. A gravity separation prediction model was built and the average fitting error of the model was 2.89%. Second, the calorific value – (ash and sulfur) model was built, in which SPSS was used to eliminate abnormal points. A mathematical model of the price of clean coal – calorific value was built. Finally, a genetic algorithm was used to optimize each sorting process to find the best sorting density of each sorting system: the result of the shallow slot heavy medium separator was 1.72 g/cm³, that of the heavy medium cyclone was 1.99 g/cm³, and the result of the spiral separator was 1.86 g/cm³. The maximum sales revenue of raw coal per ton from the economic benefit maximization prediction model was 379.79 yuan. The models provided forward-looking guidance for the related adjustments of production.