基于声发射的煤与瓦斯突出预测研究-控制理论与控制工程专业论文.docx

- - I - 摘 要 近年来，我国煤矿安全事故频频发生。其中煤与瓦斯突出现象，给煤矿安全生 产，特别是井下工作人员的生命财产造成了极其严重的威胁。这给我国煤炭行业的 发展敲响了警钟。煤与瓦斯突出是矿井含瓦斯煤岩体呈粉碎状态从煤岩层中向采掘 空间急剧(数 秒钟到数分钟完成)运 动并伴随着大量瓦斯喷出的一种强烈的动力过 程。长期的防治工作和科学研究表明，煤与瓦斯突出虽然是突发性的，但在突出前 均有前兆显现，其中声发射就是前兆之一。由于声发射信号中含有大量的噪声，而 且目前降噪技术也不能很好的使声发射监测得以准确预报，因此本论文采用了基于 蚁群优化的提升小波包渐变式阀值的降噪方法，并通过大量的数值计算和计算机仿 真证明了该思想的可行性。 提出了基于蚁群优化的小波神经网络煤与瓦斯突出的预测模型，系统研究了煤 与瓦斯突出声发射的神经网络预测的原理，将声发射自适应神经网络模型应用于煤 与瓦斯突出危险性预测上。实现了煤与瓦斯突出危险性的声发射动态趋势预测。文 章的最后还简要的介绍了监测系统的设计实施方案。 关键词：煤与瓦斯突出；声发射；小波包渐变阀值；神经网络；蚁群优化 - - II - Abstract In recent years, Chinas coal mine accidents occur frequently.Including coal and gas Outburst phenomenon,to the coal mine safety production,especially the staff of the underground life and property caused by an extremely serious threat.This gives the development of Chinas coal industry has sounded the alarm.Coal and gas outburst is the gas containing coal mine rock was crushed to the state from coal mining space rocks sharp(a few seconds to several minutes to complete)along with a lot of movement and emitted a strong gas outburst,although unexpected,but the former are precursors of the highlights show,which is the precursor of electro magnetic radiation.As the scoustic emission signal contains a lot of noise,and noise reduction technology can not currently make a good prediction based on gradient type lifting wavelet packet threshold of noise reduction method,and through a large number of numerical calculations and computer simulations demonstrate the feasibility of the idea. Proposed based on wavelet neural network ant colony optimization of coal and gas outburst prediction model,systematic study of the acoustic emission of coal and gas outburst prediction of neural network theory,the acoustic emission adaptive neural network model is applied to coal and gas outburst prediction.To achieve a coal and gas outburst prediction of the dynamic trend of acoustic emission.Finally,the article briefly describes the monitoring symtem design implementation. Key Words ： coal and gas