【作者】 苏秀平；

【导师】 李威；

【作者基本信息】 中国矿业大学 ， 机械设计及理论， 2013， 博士

【摘要】 采煤机是综采工作面的关键设备之一，是实现高效集约化采煤、减少煤矿重大恶性事故和改善工作面劳动条件的重要技术装备。目前，综采工作面技术研究重点是采煤机、液压支架和刮板输送机的三机协调控制技术，即所谓“三机联动”技术，这是为今后实现综采工作面全自动无人工作所进行的必要前期研究。当前，液压支架和刮板输送机已经实现联动自动化控制，而采煤机的单机自动控制还没有真正实现，更不用说实现与液压支架和刮板输送机的三机联动控制了。因此，采煤机的自动化控制已成为实现整个工作面自动化控制的关键。其原因是实现采煤机自动控制的瓶颈技术——滚筒自动调高控制及其有关关键技术，至今仍没有非常完善的解决方案。本论文结合国家“863”计划重点项目——“采煤机远程控制技术及监测系统”（编号：2008AA062202）的实施，在采煤机记忆截割控制技术基础上，围绕实现采煤机自动调高控制的关键技术进行研究。具体来说，本论文主要研究内容如下：（1）分析了进行采煤机自动调高控制技术研究需要解决的关键问题。通过对采煤机自动调高系统的组成和特点分析，指出采煤机调高机构截割负载的动态模拟、采煤机煤岩界面截割模式自动识别方法和采煤机自动调高技术控制策略研究是进行采煤机自动调高控制技术研究的关键问题，并分析了这些关键问题的研究进展情况。（2）对具有记忆截割功能的采煤机在工作面现场人工示教试验阶段拾取的截割电流等截割力响应信号进行分析。在对煤岩介质物理机械性质进行了解和总结前人对螺旋滚筒截割负载理论研究成果的基础上，对工作面现场人工示教阶段测试得到的截割电机电流及牵引电机转矩的分布进行分析，分析结果表明，采煤机截割力响应信号能反映截割介质的本质差别。在此基础上，进一步对滚筒截割负载的分布进行了研究，为通过可编程实现连续模拟各种煤岩截割介质及后续有关实验研究提供了参数条件。（3）利用模糊RBF神经网络对记忆截割条件下的煤岩界面截割模式自动识别算法做进一步深入研究。根据采煤工艺、当前采煤机强力截割实际和实际截割经验，改进采煤机煤岩界面截割模式自动识别算法，研究模糊RBF神经网络对所有的煤岩界面截割模式，包括正常截割、截割硬煤层、截割顶板、截割夹矸和截割断层等截割模式自动识别中的算法及应用。（4）采煤机自动调高液压控制系统的滑模控制研究。主要包括根据采煤机调高系统状态方程设计滑模面切换函数和滑模控制规律，得到滑模控制函数。在建立采煤机液压调高控制系统状态方程和滑模控制函数的基础上，进行采煤机调高控制器仿真分析。（5）采煤机自动调高控制关键技术实验研究。包括利用模糊RBF神经网络进行煤岩界面截割模式自动识别的可行性，以及利用滑模控制策略来控制采煤机滚筒进行目标截割路径跟踪时的稳定性、准确性与快速性研究。对工作面现场截割电机电流、牵引电机转矩采样值数据特征和滚筒截割负载数据特征的分析表明：采煤机截割力响应信号的均值特征能真实反映滚筒的截割工况。在本截割实验中，与纯煤截割时相比，滚筒截割底板时截割电机电流均值增加了48．1%，截割顶板时截割电机电流均值增大了8.7%；采煤机实际工作过程中，在各典型截割工况下，螺旋滚筒截割力响应信号如截割电流、牵引转矩等参数服从正态分布；在截割底板、正常截割和截割顶板工况下，正常截割工况下的截割力响应信号的方差最小，说明正常截割工况下截割力响应信号波动最小；截割状态持续时间是滚筒截割状态识别的重要参数之一；根据截割力响应信号的特征可以推出，采煤机在各截割工况下，滚筒上主要截割负载如截割阻力、截割阻力矩等服从正态分布。在截割底板、正常截割和截割顶板工况下，正常截割工况下的截割阻力、截割阻力矩的方差最小，即正常截割工况下载荷波动最小。实验研究表明，把改进的煤岩界面截割模式自动识别算法用于训练、测试模糊RBF神经网络成功后，将其用于煤岩界面截割模式识别，其输出的滚筒截割模式识别结果与滚筒实际截割模式相同，该模糊RBF神经网络能对所有的煤岩界面截割模式，包括正常截割、截割硬煤层、截割顶板、截割夹矸和截割断层等截割模式实现自动识别；同时，实验研究结果还表明，基于滑模控制的采煤机滚筒在整个目标截割路径跟踪过程中，最大跟踪误差为6mm，能满足采煤机调高要求。更多还原

【Abstract】 A shearer is one of the key equipment of mechanized mining face, and is animportant technical equipment for achieving efficiently intensive coal-mining,reducing the fatal accidents of coalmine and improving the working conditions ofworking face. At present, the research focus of the fully mechanized coalfacetechnology is the three machines coordination control technology of a shearers, ahydraulic support and a scraper conveyor,, i.e., the so-called “three-machineinteraction” technology, which is necessarily preliminary studies for the futuralrealization of automatic unattended for fully mechanized coal face.The linkageautomatic control of hydraulic support and scraper conveyor has been achieved, butstand-alone automatic control of a shearer has not really realize, let alone linkagecontrol with hydraulic support and scraper conveyor. Therefore, the shearerautomation control has become the key to the entire face automatic control. Thereason is the bottleneck technology for shearer automatic control---The drumautomatical height control and its related key technology has still not a perfectsolution.Connection with the state “863” project—“shearer remote control andmonitoring systems”(No:2008AA062202), the paper studies the key technologiesfor shearer automatic height control. Specifically, the main contents of the paper areas follows:(1) Key issues needed to be resolved for sheaere automatic control were analyzed.Through the analysis of the composition and characteristics of the shearer automaticheight control system, the paper pointed out that the dynamic simulation of cuttingloads, coal-rock interface cutting mode automatic identification method and controlstrategy research for automatic height adjustment were key technologies to carry outshearer automatic height control research.(2) Cutting current and other cutting force response signals detected in themanual teach stage at coal face are analyzed,and dynamic simulation expert systemfor cutting load are designed. On the basis of understanding physical and mechanicalproperties of coal and rock and summarizing predecessors’ theoretical research aboutthe cutting load of spiral drum, analysis of digital characteristics and distribution ofcutting current traction torque detected at face testing ground was carried out, and thedistribution of drum cutting loads was further studied. So the continuous simulation of the distribution of the various coal and rock media was realized programmablely.(3) Automatic recognition algorithm for cutting mode of the coal-rock interfacewas furtherly studied using fuzzyRBFneural network under memory cuttingcondition. According to the mining process, powerful cutting feature and actualcutting experience, the coal-rock interface automatic recognition algorithm wasimproved, and the application of FuzzyRBFneural network in the recognition of allcutting modes was researched,(4) The application of sliding mode control in the shearer automatic heighthydraulic control system was Indagated. Switching function of the sliding surface andsliding mode control law were designed and the sliding mode control function wasobtained based on the state equation of the shearer automatic lifting control system.On the basis, the sliding mode control of the shearer automatic lifting hydraulicsystem was simulated and experimentally verified.(5) Experimental studies of key technologies for shearer automatic height control.It mainly includes the feasibility study using fuzzy RBF neural network forautomatic identification of coal-rock interface cutting mode and the stability, accuracyand rapidity research using sliding mode control strategy for the drum target cuttingpath tracking.Experimental studies showed that the output cutting pattern recognition results offuzzy RBF neural network after successful training and testing is the same as theactual cutting mode of the drum. Hence, it is feasible that using fuzzy RBF neural networkfor automatic identification of the shearer drum coal-rock interface cutting mode; in theoverall objective of the cutting path tracking, the maximum tracking error of theshearer, based on sliding mode control, was6mm and could meet the requirements ofshearer height control.更多还原