基于双算法的钻机智能防卡控制

Intelligent control method of anti-sticking for drilling rod based on double algorithms

  • 摘要: 松软煤层钻孔过程中由于瓦斯含量高,易发生卡钻事故。为解决松软煤层钻孔施工卡钻问题,综合应用钻孔排渣理论和液压防卡钻技术,基于模糊神经网络和粒子群算法,提出一种智能防卡钻方法,并设计智能防卡控制器实现智能防卡控制。智能防卡钻方法,以钻杆的回转扭矩、回转速度和单位时间排渣量为信号,判断钻进状态,通过控制钻杆的给进速度进而控制煤渣的产生速率,通过控制钻杆的回转速度控制煤渣的排出速率,使孔内的煤渣产生量小于钻杆的最大排渣量,防止孔内出现煤渣堵塞,避免卡钻。智能防卡控制器由智能决策器和执行器两部分组成,智能决策器采用模糊神经网络构建模型,并使用粒子群算法进行求解,分别采用扭矩传感器、转速传感器及重量传感器实时监测钻孔施工中的钻杆扭矩、钻杆转速及排渣量,并将数据输入至智能决策器,智能决策器根据防卡钻方法输出控制决策至执行器,执行器是基于粒子群算法的最优PID控制器,控制液压机构完成智能决策器输出的决策值。仿真结果表明,模糊神经网络和粒子群算法结合与经典的误差反馈传播式模糊神经网络相比,误差小,求解速度快。智能防卡控制器可预测卡钻,并自行作出决策,防止发生卡钻事故。在液压防卡钻基础上引入钻孔排渣理论并与智能算法结合提出智能防卡控制,为预测和防止卡钻提供了理论支撑和新思路。

     

    Abstract: Drilling soft coal seam with high gas content easily leads to a sticking of drilling rod.In order to solve the above problem, an intelligent anti-sticking drilling method is put forward based on the theory of slag discharge, hydraulic anti-sticking drilling technology, fuzzy neural network, and particle swarm optimization(PSO).The intelligent anti-sticking controller is designed and applied to a proposed method which gathers the signals of torque, rotary speed, and slag discharge of drilling rod to distinguish drilling status.The generation rate and the discharge rate are controlled by the feeding speed and the rotary speed of drilling rod respectively, so that the production amount of coal particles is less than the maximum discharge amount of coal particles, thus coal particles blockage and sticking of drilling rod can be prevented.Intelligent anti-sticking controller is composed of intelligent decision-maker and executor, and the former is constructed based on fuzzy neural network.Particle swarm optimization is used to solve the model.Torque sensor, rotation speed sensor, and weight sensor are used to detect torque of drilling rod, rotary speed of drilling rod, and slag discharge in real time respectively, and the data are fed into the intelligent decision-maker.Next the control decision is transferred to the executor according to the anti-sticking drilling method.The executor is an optimal PID(Proportion Integration Differentiation) controller based on PSO,which control the hydraulic executors to complete the decision of the intelligent decision-maker.The simulation results show that the double algorithms of fuzzy neural network and PSO have achieved a smaller error and faster solving speed than the classical error feedback propagated fuzzy neural network, and the intelligent anti-sticking controller can predict the sticking of drilling rod and make the corresponding decision automatically.The intelligent control method of anti-sticking for drilling rod provides a theoretical support and a new idea for preventing the sticking of drilling rod.

     

/

返回文章
返回