支架自动跟机模式下的泵站启停控制自主决策模型的研究

Study on intelligent start-stop control strategy of emulsion pumps under the mode of support automatic follow-up

针对综采工作面自动化生产过程中液压支架跟机速率慢的问题,分析了液压支架动作与供液系统输出流量的关系以及跟机移架速度与采煤机牵引速度匹配的关系,研究了采煤机牵引速度、位置、方向与液压支架动作以及供液系统出口压力相互之间的约束关系,提出一种利用受控系统的I/O数据来设计乳化液泵启停决策模型的方法;使用实际生产数据训练了CART随机森林分类树决策模型,并针对数据不平衡问题提出一种调整权重的分类树优化算法。经实际应用数据验证,该方法不仅提高了供液系统的流量按需供给度,而且具备较好的普适性,可以应用于不同的自动化综采工作面,为提高综采工作面智能化程度以及高效生产提供支撑。

Aiming at the problem of slow follow-up speed of hydraulic support in the automatic production process of fully mechanized mining face,this paper analyses the relationship between the action of hydraulic support and the output flow of hydraulic supply system,and the matching relationship between the moving speed of hydraulic support and the speed of shearer,and studies the restriction relationship between the speed,position,direction of shearer and the action of hydraulic support and the outlet pressure of hydraulic supply system.A method of designing controller based on I/O data of controlled system is proposed.A CART random forest classification tree model is trained with actual production data to control the emulsion pump.The practical application results show that the method not only improves the flow on demand supply degree of the fluid supply system,but also has good universality and can be applied to different fully mechanized automatic mining faces.It provides support for improving intelligent degree and efficient production of the fully mechanized mining face.