基于EEMD和LSSVM的钢丝绳输送带早期故障诊断研究

发布时间：2018-10-08 21:57

【摘要】：随着科技的发展，钢丝绳输送带因其负载重、运输量大、传输距离长等特点，成为煤矿、钢铁和冶金等行业的主要运输设备，并且这些行业对钢丝绳输送带的依赖程度日益增强，人们也越来越关注钢丝绳输送带的运行状态。处于长期运行中的钢丝绳输送带，一旦发生断丝、形变、磨损等故障，会造成钢丝绳的强度下降直至断裂，最终造成严重的人员伤亡和经济损失。为了降低煤矿事故的发生，本文陈述了钢丝绳输送带无损检测国内外的发展现状，分析了钢丝绳输送带故障的原理和金属磁记忆检测的机理，研究了金属磁记忆信号的降噪算法，证明了最小二乘支持向量机理论在钢丝绳输送带早期故障诊断中的先进性和可行性。首先，介绍了钢丝绳输送带无损检测技术的发展现状和金属磁记忆技术的研究现状。通过分析钢丝绳输送带故障产生的原因，研究了金属磁记忆技术的作用机理及金属磁记忆法在故障检测中的优势。与传统检测方法对比，提出金属磁记忆技术应用于钢丝绳输送带检测，并对其进行了可行性分析。其次，金属磁记忆信号非常微弱，极易受到现场环境的干扰，如果不进行降噪处理会严重影响检测结果。依据集合经验模态分解在信号处理领域的突出特点，提出了改进型的集合经验模态分解法对金属磁记忆信号进行降噪。通过集合经验模态分解法与金属磁记忆技术相结合，可以准确的判定钢丝绳输送带应力集中的区域。然后，，从降噪的金属磁记忆信号中提取多个特征量，输入到最小二乘支持向量机早期故障诊断系统内。基于建立的早期故障诊断系统，识别和诊断钢丝绳输送带的运行状态。最后，选用粒子群优化算法对最小二乘支持向量机的参数进行寻优。仿真结果表明，该早期故障诊断系统能够实现对钢丝绳输送带状态的识别，具有较理想的准确性。
[Abstract]:With the development of science and technology, steel rope conveyor belt has become the main transportation equipment in coal mine, steel and metallurgical industry because of its heavy load, large transport capacity and long transmission distance. And these industries rely more and more on steel rope conveyor belt, people pay more and more attention to the running state of steel rope conveyor belt. The wire rope conveyor belt in long-term operation, once broken wire, deformation, wear and other failures, will cause the strength of the wire rope down to fracture, resulting in serious casualties and economic losses. In order to reduce the occurrence of coal mine accidents, this paper describes the development status of non-destructive testing of steel rope conveyor belt at home and abroad, analyzes the principle of wire rope conveyor belt fault and the mechanism of metal magnetic memory detection. The de-noising algorithm of metal magnetic memory signal is studied. It is proved that the least square support vector machine theory is advanced and feasible in the early fault diagnosis of steel rope conveyor belt. Firstly, the development of nondestructive testing technology of steel rope conveyor belt and the research status of metal magnetic memory technology are introduced. Based on the analysis of the causes of the fault of the steel rope conveyor belt, the mechanism of the metal magnetic memory technology and the advantages of the metal magnetic memory method in the fault detection are studied. Compared with the traditional detection method, the application of metal magnetic memory technology to the detection of steel rope conveyor belt is put forward, and the feasibility analysis is made. Secondly, the metal magnetic memory signal is very weak, so it is easy to be disturbed by the field environment. If the noise reduction is not carried out, the detection results will be seriously affected. According to the outstanding characteristics of set empirical mode decomposition in the field of signal processing, an improved set empirical mode decomposition method is proposed to reduce the noise of metal magnetic memory signal. Through the combination of empirical mode decomposition method and metal magnetic memory technology, the area of stress concentration of steel rope conveyor belt can be accurately determined. Then, several features are extracted from the noise-reducing metal magnetic memory signal and input into the least squares support vector machine (LS-SVM) early fault diagnosis system. Based on the established early fault diagnosis system, the running state of steel rope conveyor belt is identified and diagnosed. Finally, the particle swarm optimization algorithm is used to optimize the parameters of least squares support vector machine. The simulation results show that the early fault diagnosis system can recognize the state of steel rope conveyor belt and has a better accuracy.
【学位授予单位】：太原理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TD50

【参考文献】