综采工作面推移油缸通过电液换向阀控制推移油缸运动,常出现切换延时、控制精度差、直线度差等问题,介绍了推移油缸工作原理,建立了推移油缸运动方程,设计出推移油缸定位系统。通过RBF神经网络方式对推移油缸活塞杆t1-t3两个阶段进行精准预测,推导得出实际输出值与目标值的误差函数关系。通过Matlab仿真将目标位置导入,仿真位移与目标位置进行对比,曲线拟合程度较高。经过王坪煤业试验验证,按照控制试验原理图进行连接,得出实际位置与目标位置对比图,两者位移存在3~5mm误差,拟合程度较高,实现了精确定位控制的目的。

The movement of the hydraulic cylinder in the fully mechanized mining face was controlled by anelectro-hydraulic directional valve, which often leads to problems such as switching delay, poor control accuracy,and poor straightness. The working principle of the hydraulic cylinder was introduced, and the motion equation ofthe hydraulic cylinder was established. A positioning system for the hydraulic cylinder was designed. By using RBFneural network to accurately predict the two stages of pushing the piston rod of the oil cylinder, the error functionrelationship between the actual output value and the target value was derived. Importing the target position throughMatlab simulation and comparing the simulated displacement with the target position shows a high degree of curvefitting. After experimental verification by Wangping Coal Industry, the actual position and target positioncomparison diagram were obtained by connecting according to the control test principle diagram.There is adisplacement error of 3 ~ 5 mm between the two, and the fitting degree is high, achieving the goal of precisepositioning control.