【作者】 左红艳；

【导师】 罗周全；

【作者基本信息】 中南大学 ， 安全管理工程， 2012， 博士

【摘要】 随着中国入世以来的经济全球化的飞速发展,中国金属矿资源市场已不可避免地融入金属矿资源国际市场。世界金属矿资源市场的需求变化与价格波动及国际金属矿资源开采形势的发展对中国金属矿资源开采企业的影响十分巨大。因此,中国金属矿资源企业在市场需求与价格竞争等方面面临着前所未有的非常严峻的挑战。如何有效地减小因安全事故带来损失,降低安全管理费用和生产成本,是中国金属矿资源企业提高金属矿资源开采的盈利水平的十分重要的关键因素。而多学科设计优化方法、模糊理论、人因工程、熵理论、时滞微分方程理论、神经元网络理论等非线性理论与人工智能方法在有效地减小因安全事故带来损失,降低安全管理费用和生产成本方面具有明显的优势,可望为中国金属矿资源企业开采盈利水平提高提供理论基础与技术保障。为此,论文的研究以国家“十一五”科技支撑计划课题—金属矿大范围隐患空区调查及事故辨识关键技术研究[2007BAK22B04-12]和国家自然科学基金项目—金属矿山采场冒顶声发射信号混沌辨析及其智能预报研究[51274250]为依托,重点开展地下金属矿山开采安全生产规模、地下金属矿山开采人机安全非线性动态演化机理、地下金属矿山开采过程人机环境系统安全熵、地下金属矿山开采多因素耦合灾害智能预测与地下金属矿山开采灾害征兆参数时间序列智能预测的研究,取得主要研究成果与创新如下：(1)首次以地下金属矿山开采过程中的生产收益、安全度和环境影响度为子系统目标函数,以自适应变尺度混沌免疫优化算法为手段构建了地下金属矿山开采安全生产规模辨析模型,并对地下金属矿山开采安全生产规模进行了有效辨析和整体配置优化研究,实用结果表明,某地下铅锌矿生产规模约为125万t/a,经济寿命为14a,对应的生产收益系数可增加15.13%、安全度系数可增加5.4%和环境影响度系数可降低9.52%；(2)考虑马虎水平对安全水平的影响时滞性,构建了地下金属矿山开采人机系统马虎性与安全性的非线性动态演化模型,其有效性较好地得到仿真结果的验证,通过对地下金属矿山开采人机安全非线性动态演化及其趋势定性分析成功地揭示了动态区域内马虎水平与安全水平的相互作用演化模式,为地下金属矿山开采人机系统的安全性评价与控制提供了依据：(3)采用耗散结构理论和熵变方程建立了地下金属矿山开采人机环境系统安全熵阈限模型,揭示了初始总安全熵过高时地下金属矿山开采人机环境系统容易崩溃的机理,并构建了地下金属矿山开采人机环境系统安全熵辨析模型,应用结果表明,影响地下金属矿山开采人机环境系统安全熵辨析值的主要原因是由于该人机环境系统安全没有协调发展,导致熵产的产生；(4)在采用聚类方法确定径向基函数参数的基础上,构建了基于自适应变权重模糊RBF神经网络的地下金属矿山开采多因素耦合灾害智能预测模型,并对其进行了训练、检验和实际应用,结果表明：该预测模型精度高且泛化能力较强,并在以南方某地下金属矿山为例进行大规模采空区围岩失稳预测中得到很好验证；(5)采用将原始输入时间序列的单个预测模型预测值按正交的三角函数扩展的方法建立了函数链神经网络预测模型,结合模糊自适应变权重算法计算函数链神经网络权重,提出了基于模糊自适应变权重算法的函数链神经网络的地下金属矿山开采灾害征兆参数时间序列智能预测方法,并成功应用于某铅锌矿冒顶预报,应用结果表明,该智能预测模型需要建模数据少,能满足非线性的预测要求,并具有较高的预测精度和好的泛化能力,为金属矿山采场冒顶准确预测提供了新方法。更多还原

【Abstract】 The metal market of China has entered into international market since China became one of WTO and rapid devolepment of economic globalization, and influences from changes of metal mineral resources market the world and development of international metal mineral resources mining are enormous, therefore, the metal mineral resources industry in China will encounter a very flinty challenge. Obviously, it is a most vital and urgently solved problem for the metal mineral resources industry how to reduce accident loss, how to lower safety management costs and production costs and how to improve the economy benefits and competition capability in the metal mineral resources exploitation.But nonlinear science theory and artificial intelligent method including multidisciplinary design optimization, human factors engineering, the theory of differential delay equations, the theory of neural network, the theory of entropy and the theory of fuzzy are of unique superiority in solving the mentioned problems and will be able to offer a new method for solving them.The main research task in this paper such as safety production scale model of underground metal mine, evolution mechanism of man machine safe nonlinear dynamic for mining mineral resources in the underground metal mine, instability identification about large scale underground mined-out area in the metal mine, safety entropy analysis on man machine environment system for mining mineral resources in the underground metal mine, multi factor coupling disaster prediction in the underground metal mine and disaster evidence parameter time series prediction in the underground metal mine were studied. And the research fund is obtained from the national "eleven five" science and technology support program(namely survey on metal mine area hidden cavity and research on the key technologies of fault identification[2007BAK22B04-12]) and the National Natural Science Fund Projects(namely study on chaos discrimination for acoustic emission signals from roof caving in metal mine and its intelligent forecasting[51274250]), the main innovations and achievements of the paper are expressed as:(1)An analysis model of safety production scale of underground metal mine was established by using of some methods such as making income of production, safety and environmental impact as the subsystem objective function and using adaptive mutative scale chaos immunization optimization algorithm to solve multidisciplinary design optimization model, and safety production scale of underground metal mine was analyzed validly and optimized overally. Practical results show that multidisciplinary design optimization on production scale of an underground lead and zinc mine is more realistically reflect the actual operating conditions, the production scale is about1.25million t/a, the economic life approximately is14a, corresponding coefficient of production profits can be increased to15.13%, safety factor can be increased to5.4%and environmental impact coefficient can be reduced by9.52%.(2)Considering time delay when sloppy level influencing on safety level, the safety nonlinear dynamic evolutionary model of man machine system for mining mineral resources in the underground metal mine was proposed based on human sloppy and safety and its validty was tested by using of simulation results. Qualitative analysis on nonlinear dynamic evolution of man machine system for mining mineral resources in the underground metal mine and its trend reveal evolutionary patterns of interaction of sloppy and safety qualitatively in four dynamic regions. And some theory basis for safety evaluation and control to the underground metal mining mechanical system is provided by the research results.(3)A safety entropy thresholds model of man machine environment system for mining mineral resources in the underground metal mine was established and analyzized qualitatively based on dissipative structure theory and entropy change equation and the results reaveled that man machine environment system for mining mineral resources in the underground metal mine will collapse when The initial total safety entropy is too high. Moreover, safety entropy analysis model of man machine environment system for mining mineral resources in the underground metal mine was established, the result shows that the main factor which affects the safety for the man machine environment system is that the safety for man machine environment system doesn’t develop coordinately and leads to entropy production into man machine environment system in the underground metal mine.(4)The parameters of radial basis function were determined through clustering method and prediction model of multi factor coupling disaster from underground metal mining based on adaptive variable weight FRBFNN model was built. The prediction model of multi factor coupling disaster from underground metal mining was trained, tested and applied. The results show that adaptive variable weight FRBFNN model has high training accuracy and generalization ability. Correctness of analysis about adaptive variable weight FRBFNN model was proved by the practical application results about instability discrimination of surrounding rock in large-scale underground mined-out area of a metal mine in south China.(5)After a functional link neural network forecasting method was established by using of the method that it made some forecasting values from different single forecasting model being extended according to orthogonal trigonometric function and the weight of functional link neural network was calculated based on fuzzy adaptive variable weight algorithm fuzzy adaptive variable weight algorithm, prediction method of disaster warning parameter time series in underground metal mining based on fuzzy adaptive variable weight function link neural network was brought up. The success forecasting results of happening rate of acoustic emission in some lead&zinc mine reveal that the functional link neural network forecasting method based on fuzzy adaptive variable weight algorithm is higher than that of other forecasting model. The functional link neural network forecasting is very useful for predict roof caving.更多还原