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

The project remains with a multidisciplinary crossed and nonlinear science applied theory research fields, which combines with the mining engineering, safety engineering, detection and information engineering and artificial intelligence technology. For the low accuracy limitions of roof caving prediction process based on the deterministic mechanics model or experience model, improved phase space reconstruction theory, and chaos similar guideline which will be used to solve the problem that not easy measuring weak signal chaos identification must amplify the signal firstly but need keep the characteristics of the original signal unchanged, will be put forward in this project, and can be used to do chaos identification of the dynamic sysytem which acoustic emission parameters are changing during the metal mine rock roof caving process. Based on the chaos mechanism identification, the combined method of chaos theory, catastrophe theory, artificial intelligence technology is adopted to predict the acoustic emission parameters of metal mine rock roof caving in the time domain, and some foundation researches of metal mine rock roof caving mechanism are carried out, which can provide very good theory supports for making clear metal mine rock roof caving reckoning mechanism and roof caving accurate prediction. Project research results, with respect to revealling metal mine rock roof caving happening mechanism, not only the related theorys can reach the international advanced level, but also have significant practical value.

本项目属于采矿工程、安全工程、检测与信息工程以及人工智能技术相结合的交叉性非线性科学应用理论研究范畴。针对建立于确定性的力学模型或经验模型下的岩石冒顶预测过程准确性不高的局限性，利用本项目将提出的改进重构相空间理论，以及能解决不易测量微弱信号混沌识别需先信号放大且保持原信号的特性不变的混沌相似准则等理论方法，对金属矿山冒顶过程声发射参数随时间变化的动力学系统进行混沌识别，并在混沌机制辨识基础上采用混沌理论、灾变理论和人工智能技术相结合的方法在时域内对金属矿山岩石冒顶声发射参数进行预测，并开展金属矿山岩石冒顶灾变机理等基础性研究，为明晰金属矿山岩石冒顶灾变机理以及冒顶精确预测预报提供很好的理论支持。项目研究成果在揭示金属矿山岩石冒顶灾变发生机制方面，不仅在理论上可达国际先进水平，同时又具有重大的实用价值。

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，是地下金属矿山提高金属矿资源开采的盈利水平的十分重要的关键因素。为此，本项目针对建立于确定性的力学模型或经验模型下的岩石冒顶预测过程准确性不高的局限性，利用本项目提出的改进重构相空间理论，以及能解决不易测量微弱信号混沌识别需先信号放大且保持原信号的特性不变的混沌相似准则等理论方法，对金属矿山冒顶过程声发射参数随时间变化的动力学系统进行混沌识别，并在混沌机制辨识基础上采用混沌理论、灾变理论和人工智能技术相结合的方法在时域内对金属矿山岩石冒顶声发射参数进行预测，并开展金属矿山岩石冒顶灾变机理等基础性研究。. 目前本项目组已出版专著1部，在《Mathematical Problems in Engineering》、《Transactions of Nonferrous Metals Society of China》和《Journal of Central South University》等期刊上发表SCI论文4篇，在《International Journal of Mining Science and Technology》等国内知名学术期刊上发表EI收录论文9篇。此外，获2016年中国有色金属工业科学技术奖一等奖1次，申请发明专利1项，培养博士后1名，培养博士生2名，培养硕士生1名。. 项目研究成果在明晰金属矿山岩石冒顶灾变机理以及冒顶精确预测预报方面具有重要的理论与应用价值。