【作者】 范京道；

【导师】 常心坦；

【作者基本信息】 西安科技大学 ， 安全技术及工程， 2013， 博士

【摘要】 井巷风阻是矿井通风网络解算的重要基础数据，精度要求高，现场测取费工费时，直接影响到网络解算技术在现场的应用。风速/风量监测属于矿井安全监控系统的重要监测功能。深入提取风量监测数据中隐含的包括风阻波动变化在内的系统信息，能够形成对矿井通风计算常态化应用的有力支撑，对于提高矿井通风安全管理的精细化水平具有重要的意义。常规操作的通风网络解算在验证和认定风阻数据时需要基于经验的人为介入，无严格的程序可遵循，难以准确反映矿井客观实际。有鉴于此，本论文在综合有关风量和风阻分布数据全部信息的基础之上，通过优化平衡实现了风量监测数据的全局性推演，进而以得到验证的全矿风量分配格局为目标和限制条件，获取了各风道风阻数据的数学模型，并设计了风量和风阻计算的两级控制和校验，有效改善了模型的适用性及可靠性，显著降低了对人为介入的要求，并稳定提高了计算数据的客观符合度，能够形成对通风网络解算技术在煤矿现场常态化应用的有力支撑。在论文研究中，提出了将日常生产条件下的监测风量变化分为风量波动和风量漂移的概念；提出了风量波动属于白噪声平稳序列，而风量漂移属于非平稳序列的论点。在此基础上以泰勒级数展开将风量变化的非线性函数线性化，构建了风量波动与漂移溯源分析的数学模型，以通风网络的灵敏度矩阵为指导，辨识风量变化的致因分支及其风阻变化量，进而获得矿井风量/风阻波动的统计特征及风阻漂移数据。为定量描述属于非线性复杂动态系统的通风系统变化特征迈出了重要的一步，可有效指导矿井通风的隐患辨识与风阻数据维护。论文研究表明，风量变化溯源分析在整个通风系统内的辨识力存在一定的区域性差异，因而影响辨识精度的均匀性，甚至形成局部的“辨识盲区”而引起溯源分析的多解问题。论文研究因此指出了通风系统内风量监测点布局优化的必要性，提出了基于单风道风阻变化影响分析的布局优化和设置附加判据处理多解辨识的方法，提高了溯源分析对网络系统的适应能力及辨识精度。本论文研究提出了通风系统数字实验平台的概念，基于随机过程和蒙特卡罗方法的有关理论提供了完整的数字实验设计，并以数字实验验证了风量波动与漂移溯源分析的数学模型。通风数字实验平台能够提供高度接近现场通风实际的模拟场景，适宜于在广泛的研究范围内展开专题论证与方案分析，具有明显的应用前景。更多还原

【Abstract】 As basic data of mine ventilation network calculation, resistances of airways needs to bemeasured accurately through time-consuming work process, which seriously hinders theapplication of ventilation network calculation in mines. Being one of the important functionscovered by Safety Monitoring Systems, systematically operated airflow measurement canreveal a wide range of ventilation information such as changes of resistance distribution.Relevant analysis of airflow data can provide valuable support to the routine application ofmine ventilation calculations, thus promote mine safety management to a higher level.As a tool for engineering problem solving, mine ventilation network calculationsnormally do not have sufficient and accurate data to balance the network while allow theresult fit the in-situ measurement accurately, thus, when applying to a mine, ventilationnetwork solving pursues a rational rather than theoretically accurate solution, and is lack ofrigid rule to follow as well as a clear and definite objective to pursue. This dissertation putsforward a mathematical model to deduce airflow pattern in the system based on limitedmonitoring data of airflow, and further aim at the deduced airflow distribution data to obtainairway resistance data in an optimization process. A two-phase control design is adopted toensure its applicability and reliability with a lower need of operation interference and a higherdegree of rationality, thus opens a door for the technique of ventilation network calculationinto daily safety management.In the dissertation, two concepts of airflow fluctuation and airflow drift are put forwardto mark off airflow variances of different nature. It is pointed out that airflow fluctuationshave the characteristics of white noise sequence while airflow drifts demonstrate thecharacteristics of gradual change sequence. Based on the application of Taylor Series and relevant mathematical theories, a ventilation sensitivity matrix is calculated to aid theidentification of ventilation resistance changes that led the observed airflow changes in thesystem, and the statistical characteristics of airflow and resistance changes can be obtained.The research of this paper shows that the identification capability of cause tracking ofairflow variance has some regional discrepancy within the whole ventilation system. Thiscould influence the uniformity of identification accuracy, and thus could form some localblind areas for identifying which in turn could result in the problem of multi-solutions.Therefore this paper suggests that it is necessary to optimize the distribution of airflowmonitors within the whole ventilation system, putting forward the methods to optimize theconfiguration of airflow monitors based on the analysis of single airway’s airflow change aswell as deal with the multi-solutions problem by setting additional criterions, which enhancethe application accuracy and suitability of cause tracking in a ventilation system.A numerical experiment platform of coal mine ventilation system is designed in thedissertation to verify the proposed model of cause tracing based on relevant theories ofstochastic process and Monte-Carlo method. The numerical experiment platform ofventilation system can effectively provide simulated sceneries of real mine network, thus issuitable for a variety of analysis and verifications relating to mine ventilation, and can bewidely applied in the future.更多还原