基于Simulink的批次磨矿产物粒度组成预测研究

Study on prediction of particle size composition of batch grinding products based on Simulink

以磁铁矿为研究对象,基于批次磨矿试验结果采用B_(II)法对构建模型中的破裂分布函数与选择函数进行求解计算,借助Simulink构建总体平衡仿真模型,进行磨矿产物粒度组成预测。结果表明:混合物料的磁铁矿批次磨矿过程符合一阶磨矿动力学模型,各粒级物料的磨矿速度与其粒度大小有关,物料的破碎速率随粒度的减小而减小。磨矿产物各粒级的预测值与试验值均相近,但在预测磨矿4 min的粒度组成时,在-0.075 mm和(-0.150+0.106)mm粒级预测的绝对误差较大,分别为3.21%和3%;在(-0.425+0.300)mm粒级预测的相对误差最大,为24.21%。预测磨矿8 min的粒度组成时,在(-0.212+0.150)mm粒级预测的绝对误差和相对误差最大,依次为0.65%和14.98%。与磨矿4 min相比,磨矿8 min的绝对误差之和与相对误差之和均较小,预测精度相对较高。

The research object of this paper is magnetite.The fracture distribution function and selection function in the model is calculated by B_(II) method based on the results of batch grinding test.The overall balance simulation model is constructed by Simulink to predict particle size composition of grinding products.The results show that magnetite batch grinding process of the mixed material conforms to the first-order grinding kinetics model.The grinding speed of each particle size material is related to its size,and the crushing rate of the material decreases with the decrease of the particle size.The predicted value of each particle size of the grinding product is close to the experimental value.When predicting the particle size composition at 4 min,the absolute error of prediction at-0.075 mm and-0.15+0.106 mm particle size is relatively large,which are 3.21%and 3%.The maximum relative error of prediction at-0.425+0.3 mm particle size is 24.21%.At 8 min,the absolute and relative errors of prediction at-0.212+0.15 mm particle size are the largest,which are 0.65%and 14.98%.Compared with 4 min,the sum of absolute errors and the sum of relative errors of 8 min are both smaller,and the prediction accuracy is relatively high.