基于声发射的煤与瓦斯突出预警与源定位研究

发布时间：2018-06-27 01:09

本文选题：声发射 + 煤与瓦斯突出　；参考：《辽宁工程技术大学》2017年硕士论文

【摘要】：煤与瓦斯突出破坏性巨大,其发生往往具有突发性和不易侦测性的特点,只能被动地预防。煤与瓦斯突出灾害发生前往往伴随着声发射(AE)现象的产生,这可以作为突出灾害预测(突出可能性以及突出点)的重要依据。实际工程及应用中引入声发射相对强弱指标(CR)进行煤与瓦斯突出预警。为了提高声发射技术在煤与瓦斯突出灾害的预警过程中的准确性,针对煤矿井下噪声来源及其性质,提出了具有针对性的声发射复合滤波方法,首先采用小波分解滤除规则干扰,然后采用中值滤波法滤除脉冲噪声,最后用奇异值分解法滤除随机噪声。为了对瓦斯突出灾害在发生发展过程中的演变过程进行充分掌握,为井下工作人员提供可能的煤与瓦斯突出位置,提出煤岩体声发射源定位方法,将优化的改进型多输出支持向量机的智能算法应用于具有特殊波形的井下声源定位,采用KPCA-MA-MSVM方法针对输入参数信号可能存在的耦合和冗余问题,以及多输出支持向量机的内部参数选择问题进行了优化,采用核主成份分析法作用于样本输入,提取出新的独立的重要影响因素,用文化基因算法作用于多输出支持向量机模型,搜索支持向量机内部最优惩罚因子C,最后,利用铅断实验来检验定位性能。结果表明:文中算法提高了试验平台声发射定位的精度且定位时间少于其他定位算法,具有很高的实际应用价值。
[Abstract]:Coal and gas outburst is very destructive, and its occurrence often has the characteristics of sudden and difficult detection. It can only be prevented passively. The occurrence of acoustic emission (AE) is often accompanied by the occurrence of acoustic emission (AE) before the occurrence of coal and gas outburst disaster. This can be used as an important basis for the prediction of the outburst disaster (prominent possibility and outburst point). In order to improve the accuracy of coal and gas outburst, the relative intensity index (CR) is used to improve the accuracy of the acoustic emission technology in the early warning process of coal and gas outburst. In view of the source and nature of the coal mine noise, a targeted acoustic emission composite filtering method is proposed. First, the wavelet decomposition is used to filter the interference of the rules, and then the noise is filtered. In order to fully grasp the evolution process of the gas outburst disaster in the course of development and provide the possible coal and gas outburst position for the underground workers, the method of locating the acoustic emission source of the coal and rock mass is put forward, and the optimized and improved multi transmission is put forward. The intelligent algorithm of support vector machine is applied to the location of sound source with special waveform. The KPCA-MA-MSVM method is used to optimize the problem of the possible coupling and redundancy of the input parameter signals and the selection of the internal parameters of the multi output support vector machine. The kernel principal component analysis is used to extract the sample input. The new independent important influence factor, using the cultural genetic algorithm to function in the multi output support vector machine model, search the optimal penalty factor C in the support vector machine. Finally, the lead breaking experiment is used to test the location performance. The results show that the algorithm improves the accuracy of the acoustic emission localization of the test platform and the positioning time is less than other location algorithms. It has high practical application value.
【学位授予单位】：辽宁工程技术大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TD713

【参考文献】