煤炭工程 ›› 2018, Vol. 50 ›› Issue (9): 117-120.doi: 10.11799/ce201809030

• 研究探讨 • 上一篇    下一篇

基于BP神经网络的液压支架支护位姿运动学分析

李海锋   

  1. 中国煤炭科工集团太原研究院有限公司
  • 收稿日期:2018-04-20 修回日期:2018-05-30 出版日期:2018-09-20 发布日期:2018-12-18
  • 通讯作者: 李海锋 E-mail:378638782@qq.com

Kinematics analysis of support position and posture of hydraulic support based on BP neural network

  • Received:2018-04-20 Revised:2018-05-30 Online:2018-09-20 Published:2018-12-18

摘要: 分析了在复杂多变的采煤工况条件下,液压支架位姿空间内变量受到驱动空间变量的耦合影响以及位姿解算应用存在的问题,设计一种基于BP神经网络的位姿空间到驱动空间的非线性转换模型,通过正向运动建立训练数据集,构造逆运动模型,能够精准快速的实现其位姿运动学计算,从而实时检测液压支架支护位姿。实验仿真结果表明,该方法控制支架位姿相对于传统方法,快速精确,提升了工作效率,为感知和调控支护位姿与跟机调控提供理论支撑。

关键词: 位姿解算, BP神经网路, 液压支架, 支护位姿

Abstract: Aalysis of the coal mining under complex conditions, Hydraulic support posture space variable by the coupling effect of driving spatial variables and pose solution of application problems, put forward using BP neural network to build the pose space to nonlinear drive space conversion model, can quickly realize the kinematics calculation, and real-time detection of hydraulic support posture. The experimental results show that the method of pose control stent compared with traditional methods, quickly and accurately, improve work efficiency, theory support for supporting a branch of perception and attitude control with machine control.

中图分类号: