集成学习在直线振动筛的应用及参数优化

针对振幅、振动方向角、振频、筛长、筛孔、筛丝直径,及筛面倾角7个因素的不同水平,共设计442组实验方案,在EDEM软件中进行模拟仿真;结合集成学习的方法对所得的实验数据进行建模和综合分析,得出不同振动参数对筛分效率的影响权重.结果表明:筛面位于波谷和波峰时,筛上颗粒平均质心位置与筛网位置的距离不一样,当差值约为1.3倍的筛孔尺寸时,其筛分效果好,且对应的触筛角度值在12°~23°浮动.

Application and Parameter Optimization of Ensemble Learning in Linear Vibrating Screen

Aiming at the different levels of amplitude, vibration direction angle, frequency, screen length, sieve pore, screen diameter and screen inclination angle, 442 groups of simulation experiments were designed using EDEM software. Combined with ensemble learning methods, the experimental data were modeled and comprehensively analyzed to obtain the affecting weights of different vibration parameters on the screening efficiency. The results showed that the distance between the average mass center of the particles on the screen and the screen was different when the screen surface was in trough and peak. When the difference between them was about 1.3 times of the sieve pore, the screening efficiency was high, and the corresponding contact angle of the screen fluctuated from 12° to 23°.