基于扰动观测器的液压锚杆钻臂自适应神经网络跟踪控制  

Adaptive Neural Network Tracking Control for Hydraulic Anchor Rod Drilling Arm Based on Disturbance Observer

作　　者：杨勇 Yang Yong(CCTEG Taiyuan Research Institute Co.,Ltd.,Taiyuan 030006,China;Shanxi Tiandi Coal Mining Machinery Co.,Ltd.,Taiyuan 030006,China)

机构地区：[1]中国煤炭科工集团太原研究院有限公司,太原030006 [2]山西天地煤机装备有限公司,太原030006

出　　处：《煤矿机械》2024年第9期171-174,共4页Coal Mine Machinery

基　　金：山西省重点研发计划项目(202102010101006);天地科技股份有限公司科技创新创业资金专项重点项目(2022-2-TD-ZD001);山西天地煤机装备有限公司青年项目(M2023-QN20)。

摘　　要：在井下复杂工况条件下,液压锚杆钻机的一些物理变量往往难以测得,且会随周围环境发生变化,这些时变参数会对控制性能产生影响。针对复杂多样的液压锚杆钻臂,首先通过状态变换得到锚杆钻臂位移系统模型空间表达式;然后采用神经网络径向基函数对未知不确定项进行逼近,并采用动态曲面技术解决计算复杂问题;最后设计了一种新型扰动观测器来估计由外部扰动和神经网络误差,结合反步法递推出控制器。根据Lyapunov函数证明系统是半全局稳定的,跟踪误差收敛至零点附近。通过与传统PID控制仿真对比,该控制器的控制精度、稳定性和鲁棒性均有显著提升。Under complex working conditions underground,some physical variables of hydraulic anchor rod drilling arm are often difficult to measure and may change with the surrounding environment.These time-varying parameters can have an impact on control performance.Aiming at the uncertain hydraulic anchor rod drilling arm system,firstly the spatial expression of the anchor rod drilling arm system model was obtained through state transformation.Then a neural network radial basis function was used to approximate the unknown uncertainty,and dynamic surface technology was introduced to solve complex computational problems.Finally,a new disturbance observer was designed to estimate external disturbances and neural network errors,and the controller was delivered by combining with backstepping method.According to the Lyapunov function,it is proven that the system is semi globally stable and the tracking error converges near zero.Compared with traditional PID control simulation,this controller has significantly improved control accuracy,stability and robustness.

关 键 词：锚杆钻臂 扰动观测器 神经网络 跟踪控制 

分 类 号：TD41[矿业工程—矿山机电]