矿用输送带损伤线性调频红外检测方法研究

The conveyor belt is the key part of the production and transportation system of the mines and metallurgical enterprises, whose fault can cause major production safety accidents. The root of the failure is the damage of the conveyor belt, whose detection and early warning basic theory needs to be improved. The project focuses on analyzing sensitivity mechanism of the electromagnetic wave echo frequency-spectrum characteristic of the conveyor belt damage. Researching on the detection method of the energy envelope and characteristic parameters (amplitude and phase) of the conveyor belt damage linear frequency modulation infrared（LFMIR） signal. Constructing the mathematical model for damage detection. Revealing the evolution law of the damage accumulation and risk behavior in the process of conveyor belt operation. Proposing the energy envelope and characteristic parameter detection method based on the acquisition and noise suppression of LFMIR damage reflection signal. Researching the power aggregation mathematical model based on LFMIR signal in the matching FRFT domain. Analyzing the coupling relationship between the power-spectrum peak parameters and the damage characteristics. Extracting damage characteristic parameters and deeply learning damage characteristics samples. Realizing the hardware in the loop simulation, performance test and risk diagnosis mechanism verification based on using the dSPACE platform to extract damage characteristics , and establishing database diagnosis theory for conveyor belt damage characteristics. The project aims to solve the technical problems of risk assessment in the process of conveyor belt operation and avoid the accidents occurrence. Meanwhile, the project establishes the theory foundation of LFMIR detection method in the area of coal mine safety detection, core equipment damage diagnosis and early warning.

输送带是矿山和冶金企业生产运输系统的关键部件，发生故障会引发重大生产安全事故，故障的根源是输送带损伤，其检测预警基础理论亟待完善。项目拟围绕输送带损伤风险电磁波回波频谱特征敏感机理，研究输送带损伤风险线性调频红外（LFMIR）反射信号能量包络及特征参量（幅值、相位）检测方法，构建损伤检测数学模型，揭示输送带运行过程损伤积累和风险行为的演化规律，提出损伤风险LFMIR反射信号采集、噪声抑制的能量包络及特征参量检测方法，研究建立在匹配FRFT域上LFMIR信号的功率聚集数学模型，分析功率谱峰参数与损伤风险特征之间的耦合关系，提取损伤风险特征参量，深度学习损伤风险特征样本；利用dSPACE平台提取损伤特征，进行半实物仿真和性能测试、风险诊断机制验证，建立输送带损伤风险特征数据库诊断理论。解决输送带运行风险评估技术难题，避免事故的发生，奠定LFMIR在煤矿安全检测、关键设备损伤风险诊断预警理论基础

输送带是矿山和冶金企业生产运输系统的关键部件。物料运输过程中，输送带容易出现纵向撕裂等损伤，且损伤一旦无法得到有效控制，就会导致输送带报废、系统设备损毁、物料洒落堵塞巷道、引发火灾、造成人员伤亡等重大生产安全事故。输送带损伤实时检测是煤矿生产中必须要解决的关键问题，研究检测精度高、抗噪性能优异、可实时检测的输送带损伤预警检测技术，实现对输送带损伤的及时发现、及时停机，可从根本上减少故障并避免事故发生。本项目围绕输送带损伤风险电磁波回波频谱特征敏感机理，研究输送带损伤风险线性调频红外反射信号能量包络及特征参量检测方法，构建损伤检测数学模型，揭示输送带运行过程损伤积累和风险行为的演化规律。结合红外成像技术和可见光成像技术的互补特性，提出一种基于红外与可见光融合的输送带损伤检测方法；设计基于红外与可见光融合的一体化双目视觉传感装置；通过对红外检测技术进一步深层次透析，结合中波红外对材质识别和长波红外对温度敏感的特性，提出一种基于中波红外和长波红外的输送带损伤检测方法；研究了红外双波段红外视觉传感器的成像原理。研究提出输送带损伤特征分析提取方法，设计输送带损伤线性调频红外分析检测算法，进行半实物仿真和性能测试、风险诊断机制验证，建立输送带损伤风险特征数据库诊断理论。搭建输送带实验平台装置，验证所提方法及相关算法正确性和有效性，输送带损伤检测的平均检测准确率为90%以上。基于项目研究，共申请发明专利11项，当前授权发明专利6项，发表学术论文15篇，其中SCI检索13篇，培养硕士研究生16人。项目研究成果为输送带平稳高效生产运行和安全决策提供有力技术辅助保障，在矿井信息化和安全检测方面有广阔应用推广前景。