【作者】 吴杰；

【导师】 袁亮；

【作者基本信息】 安徽理工大学 ， 安全工程（专业学位）， 2025， 硕士

【摘要】 在煤矿生产过程中,采掘、运输等核心环节持续产生生产性粉尘,既会危害作业人员的身体健康,也有可能会造成爆炸事故。在此背景下,研究精准高效的粉尘浓度监测技术已成为实现矿山安全生产和职业病防治的关键需求。现有基于光散射法的检测设备易受环境变化影响,且粉尘浓度较高时,光学镜头易污染;而静电感应法则存在低浓度粉尘检测灵敏度不足的缺陷。本研究地提出基于光散射法与静电感应法的粉尘浓度融合监测技术,旨在突破光散射法与静电感应法单一检测方法的局限性,实现粉尘浓度的精准检测。主要研究内容如下:(1)光散射法粉尘浓度监测技术研究。基于Mie散射理论,构建了整体光散射法颗粒物浓度的理论模型,基于自主搭建的多角度颗粒物浓度的光散射法测量实验平台,对入射光波长、探测角位置及颗粒粒径对光散射法的影响进行仿真计算研究,得到了各相关参数与散射光强的分布规律以及散射的最佳探测角度,并用实验与模拟结果验证。(2)静电感应法粉尘浓度监测技术研究。基于静电感应法基本原理推导环状电极荷电数学模型,使用Ansys maxwell模块建立了环状电极的粉尘采样管结构模型,并对环状电极的直径、长度以及厚度进行仿真分析,得到各参数的对环状电极空间灵敏度分布的影响规律,确定环状电极的最优方案为:R=25mm、L=50mm、d=2mm。(3)传感器数据融合算法及融合结构研究。为了克服光散射传感器在和静电感应传感器的缺点,选出适合当前情况的数据层融合,在此基础上选择加权平均算法、贝叶斯概率估计、卡尔曼滤波算法对传感器数据进行融合,通过Matlab比较融合结果,验证卡尔曼滤波在动态场景下的综合优势(计算耗时15.43 ms、平均误差1.206 mg/m~3,方差降幅达74.9%),采用的卡尔曼滤波算法将传感器数据在10-100mg/m~3内进行数据融合,提出多模态检测机制,在数据融合基础上对传感器融合结构进行设计,做出融合式传感器的原型。(4)粉尘浓度检测装置验证。实验测试了光散射法和静电感应法融合粉尘浓度检测装置的精确性和稳定性。在此基础上,开展了多种工况因素(环境温度、相对湿度、颗粒粒度分布及巷道风速)对传感器检测性能的影响研究,总结了不同工况因素下对传感器测试性能的影响规律。结果表明,基于融合技术的粉尘浓度传感器在10-100 mg/m~3的融合段检测误差率<10%,波动性<10%。融合传感器在融合段的平均误差相较于单一光散射法平均误差9.49%,降低了3.58%,相较于静电感应检测法平均误差13.6%,降低了32.72%。采用融合算法测量粉尘浓度的精度更高、更稳定。本研究为职业病防控和智能矿山建设提供了高可靠性的数据支撑,为矿井粉尘监测领域的精准化、智能化发展提供理论指导。图[58]表[18]参[86]更多还原

【Abstract】 During the coal mine production process,core links such as mining and transportation continuously generate high-concentration suspended dust,which not only endangers the physical health of the workers but also may cause explosion accidents.Against this background,the research on precise and efficient dust concentration monitoring technology has become a key demand for achieving safe production in mines and preventing and controlling occupational diseases.The existing detection equipment based on the light scattering method is susceptible to environmental changes.Moreover,when the dust concentration is high,the optical lenses are prone to contamination.However,the electrostatic induction law has the defect of insufficient sensitivity in detecting low-concentration dust.This study proposes a dust concentration fusion monitoring technology based on the light scattering method and the electrostatic induction method,aiming to break through the limitations of the detection methods of the light scattering method and the electrostatic induction method and achieve precise detection of dust concentration.The main research contents are as follows:Research on dust concentration monitoring technology by light scattering method based on the Mie scattering theory,a theoretical model of particulate matter concentration by the overall light scattering method was constructed.Based on the self-built multi-angle particulate matter concentration measurement experimental platform by the light scattering method,the influence of incident light wavelength,detection angle position and particle size on the light scattering method was simulated and calculated.The distribution laws of various relevant parameters and scattered light intensity,as well as the optimal detection angle of scattering,were obtained,and the experiments were verified by experimental and simulation results.(2)Research on dust concentration monitoring technology by electrostatic induction based on the basic principle of electrostatic induction method,the mathematical model of charge of the annular electrode was derived.The structural model of the dust sampling tube of the annular electrode was established using the Ansys maxwell module,and the diameter,length and thickness of the annular electrode were simulated and analyzed.The influence laws of each parameter on the spatial sensitivity distribution of the annular electrode were obtained,and the optimal scheme of the annular electrode was determined as:R=25mm,L=50mm,d=2mm.(3)Research on sensor data fusion algorithm and fusion structure:in order to overcome the shortcomings of the light scattering sensor and the electrostatic induction sensor,and select the data layer fusion suitable for the current situation,on this basis,the weighted average algorithm,Bayesian probability estimation,and Kalman filtering algorithm are selected to fuse the sensor data.The fusion results are compared through Matlab.Verify the advantages of Kalman filtering in dynamic scenes(calculation time consumption 15.43 ms,average error 1.206 mg/m~3,and variance reduction of 74.9%).The adopted Kalman filtering algorithm fuses the sensor data within the range of 10-100mg/m~3and proposes a multimodal detection mechanism.Based on data fusion,the sensor fusion structure is designed and the prototype of the fused sensor is made.(4)Verification of dust concentration detection device:The experiment tested the accuracy and stability of the dust concentration detection device that combines the light scattering method and the electrostatic induction method.On this basis,the influence of various working condition factors(environmental temperature,relative humidity,particle size distribution and roadway wind speed)on the detection performance of the sensor was studied,and the influence laws of different working condition factors on the test performance of the sensor were summarized.The results show that the dust concentration sensor based on fusion technology has a detection error rate of less than 10%and a fluctuation of less than 10%in the fusion section of 10-100 mg/m~3.The average error of the fusion sensor in the fusion section was reduced by 3.58%compared with the average error of 9.49%of the single light scattering method and by 32.72%compared with the average error of 13.6%of the electrostatic induction detection method.The accuracy and stability of measuring dust concentration by using the fusion algorithm are higher.This research provides highly reliable data support for the prevention and control of occupational diseases and the construction of intelligent mines,and offers theoretical guidance for the precise and intelligent development in the field of mine dust monitoring.Figure[58]Table[18]Reference[86]更多还原