平行掘进:基于ACP理论的掘-支-锚智能控制理论与关键技术

Parallel tunneling:Intelligent control and key technologies for tunneling,supporting and anchoring based on ACP theory

虽然智能掘进技术已取得了一些进步,提出了护盾式、掘锚式等掘-支-锚多装备的样机设计,但从设备适用性及装备配套性角度,仍存在个体装备设计不严谨、样机改型成本高、装备控制不精准、巷道场景多样不适应、群体设备难一致协同等诸多问题。造成上述问题的首要原因是缺少煤矿特殊生产环境下群体设备联动协同控制的基础理论研究,主要表现为对智能掘进所需的设备模型、场景实验、数据驱动等基础理论缺少研究。针对掘进复杂系统难以建模与实验不足等问题,提出了平行掘进系统的研究思想,试图用一种适合复杂系统的计算理论与方法解决智能化掘进现存问题,主要利用大型计算模型、预测并诱发引导复杂系统现象,通过整合人工社会、计算实验和平行系统等方法,形成新的计算研究体系。以智能掘进复杂系统为研究对象,重点考虑掘-支-锚系统的整体控制理论与方法,结合智能化掘进相关研究基础和方法,融入当前国内外复杂系统研究的最新科研进展,基于平行控制基础理论及ACP(Artificial,Computing,Parallel)方法框架,运用代理控制理念及自适应动态规划多智能体研究方法,开展智能掘进复杂系统的掘-支-锚平行控制基础理论与方法研究。针对巷道复杂环境下的群体设备离散控制现状,运用整体系统论的方法,完成掘进系统的掘进、支护、钻锚群体设备代理模型的构建,根据多场景下复杂系统的计算实验克服掘进巷道环境复杂、恶劣的工程验证难题,最终实现智能掘进平行控制方法的理论创新与方法验证。

Although intelligent drivage technology has made some progress,and the prototype design of shield type and anchor-excavation type excavation-bolt-anchor multi-equipment are presented,from the equipment applicability and equipment matching point of view,there are still many problems,such as un-rigorous individual equipment design,high cost of prototype modification,inaccurate equipment control,diverse heading scenes,un-adaptability to group equipment and difficult coordination.The primary reason for the above problems is the lack of basic theoretical research on group equipment linkage and collaborative control under the special production environment of coal mine,which is mainly manifested in the lack of basic theoretical research on equipment model,scene experiment and data drive required by intelligent drivage.The modeling of drivage complex system is difficult and the experiment is insufficient.Therefore,the research idea of parallel drivage system is put forward for the first time,trying to solve the existing problems of intelligent drivage with a kind of calculation theory and method suitable for complex system.Some large-scale computational models are used to predict and induce complex system phenomena by integrating artificial social computing experiments and parallel systems to form a new computing research system.Taking the intelligent drivage complex system as the research object,this paper focuses on the overall control theory and method of the excavation support anchor system,combines with the relevant research basis and method of intelligent drivage,integrates the latest research progress of complex system,is based on the basic theory of parallel control and ACP(artificial,computing,parallel)method framework,and applies multi-agent research method based on agent control and adaptive dynamic programming.Also,it presents the research on the parallel control theory and method of excavation-bolt-anchor for intelligent drivage complex system,current situation of discrete control of group equipment in complex environment of roadway.Based on the holistic system theory,the agent model of the drivage support group equipment of the drivage system is constructed.According to the calculation experiment of the complex system under multiple scenes,the engineering verification problem of the complex and bad environment of roadway was overcome.Finally,the theory innovation and method verification of intelligent drivage parallel control method are realized.