采煤机自适应截割轨迹预测方案与试验

Scheme and Test of Adaptive Cutting Trajectory Prediction of Shearer

针对采煤机截割控制系统存在一次截割不干净、与煤层上下边界干涉碰撞的问题,提出基于深度学习的采煤机自适应截割轨迹预测方案。在分析采煤机截割过程、轨迹预测原理的基础上,设计了具有时间、空间特性的采煤机自适应截割轨迹数据预测方案以及截割轨迹预测模型,并完成试验验证。结果表明,采煤机自适应截割轨迹预测方案能够实现对采煤机截割轨迹变化规律的预测,上截割滚筒轨迹预测均方根误差为0.022 m,下截割滚筒轨迹预测均方根误差为0.013 m,满足采煤机截割控制系统要求。

In the view of the problems of unclean one-time cutting and interference collision with the upper and lower boundaries of the coal seam in the shearer cutting control system,an adaptive cutting trajectory prediction scheme of shearer based on deep learning was proposed.On the basis of analyzing the cutting process and trajectory prediction principle of shearer,a data prediction scheme with temporal and spatial characteristics and a prediction model for adaptive cutting trajectory were designed,and the experimental verification was completed.The results showed that the adaptive cutting trajectory prediction scheme of shearer can predict the change law of shearer cutting trajectory,and the root mean square error of the upper cutting drum trajectory prediction was 0.022 m,and the root mean square error of the lower cutting drum trajectory was 0.013 m,which can meet the requirements of the shearer cutting control system.