铝电解故障诊断综合系统
发布时间：2018-08-03 06:49
【摘要】：铝电解过程是众多复杂工业流程中的一种,它具有复杂工业流程的非线性、时变和大时滞等特点。在铝电解的进行过程中,不仅存在由于电解槽内物料平衡与能量平衡等被破坏而引起的电解槽故障,还存在由于电解过程中执行器电动机的故障而造成的故障。本文中将电解槽故障定义为铝电解的过程故障,执行器电动机故障定义为系统故障。针对以上情况本文提出了铝电解多故障诊断系统,并以VB 6.0为开发平台,设计了铝电解监控系统。该多故障诊断系统包括了铝电解过程故障诊断系统和铝电解系统故障诊断系统,铝电解监控系统包括电解槽监控系统和电动机监控系统。根据以上内容本人作了以下工作。第一,阐述了铝电解故障诊断的意义,概述了铝电解故障诊断国内外的发展现状及存在的问题,针对铝电解故障诊断中还存在的问题,叙述了本课题的研究目的和研究意义,提出了自己的研究方法。第二,概述了铝电解的工作原理及铝电解的生产过程,论述了铝电解过程中几种常见的过程故障,以及铝电解过程中常见的系统故障,并对这些故障进行了简单的介绍。第三,建立了铝电解过程故障诊断模型,阐述了神经网络的基本原理,分析了神经网络在故障诊断的应用可能性和必要性,论述了铝电解过程中特征量的选取及输入数据的处理,并采用蛙跳算法对神经网络的训练过程进行了优化,仿真实验验证了铝电解过程故障诊断模型的可行性。第四,建立了铝电解系统故障诊断模型,阐述了采用EMD算法对电动机故障信号进行处理的优越性,介绍了小波分析理论及小波包分析用于故障信号特征的提取,以及小波分析与神经网络的结合,仿真实验验证了铝电解系统故障诊断模型的可行性。第五,以VB 6.0为开发平台,设计了铝电解监控软件。综述了该软件的总体思路和设计方案,并详细介绍了监控软件的各个部分功能的实现过程。第六,对本文中所采用的铝电解故障诊断方法进行了总结,并对铝电解故障诊断今后的发展做了展望。
[Abstract]:Aluminum electrolysis process is one of many complex industrial processes. It has the characteristics of nonlinear, time-varying and large time-delay of complex industrial processes. In the process of aluminum electrolysis, there is not only the failure of electrolytic cell caused by the destruction of material balance and energy balance in the electrolytic cell, but also the fault caused by the fault of actuator motor in the process of electrolysis. In this paper, the fault of electrolysis cell is defined as the process fault of aluminum electrolysis, and the fault of actuator motor is defined as the fault of system. In this paper, a multi-fault diagnosis system for aluminum electrolysis is put forward, and the monitoring system of aluminum electrolysis is designed on the platform of VB 6.0. The multi-fault diagnosis system includes aluminum electrolysis process fault diagnosis system and aluminum electrolysis system fault diagnosis system. Aluminum electrolysis monitoring system includes electrolytic cell monitoring system and motor monitoring system. According to the above, I have done the following work. First, the significance of aluminum electrolysis fault diagnosis is expounded, the development status and existing problems of aluminum electrolysis fault diagnosis at home and abroad are summarized, and the research purpose and significance of this subject are described in allusion to the existing problems in aluminum electrolysis fault diagnosis. The author puts forward his own research method. Secondly, the working principle of aluminum electrolysis and the production process of aluminum electrolysis are summarized. Several common process faults in aluminum electrolysis process and common system faults in aluminum electrolysis process are discussed, and these faults are briefly introduced. Thirdly, the fault diagnosis model of aluminum electrolysis process is established, the basic principle of neural network is expounded, and the possibility and necessity of application of neural network in fault diagnosis are analyzed. The selection of characteristic quantity and the processing of input data in aluminum electrolysis process are discussed. The training process of neural network is optimized by using leapfrog algorithm. The feasibility of fault diagnosis model of aluminum electrolysis process is verified by simulation experiment. Fourthly, the fault diagnosis model of aluminum electrolysis system is established, and the superiority of using EMD algorithm to deal with the fault signal of motor is expounded. The wavelet analysis theory and wavelet packet analysis are used to extract the feature of fault signal. Combined with wavelet analysis and neural network, the feasibility of fault diagnosis model of aluminum electrolysis system is verified by simulation experiment. Fifthly, the monitoring software of aluminum electrolysis is designed based on VB 6.0. The general idea and design scheme of the software are summarized, and the realization process of the functions of each part of the monitoring software is introduced in detail. Sixth, the methods of aluminum electrolysis fault diagnosis used in this paper are summarized, and the future development of aluminum electrolysis fault diagnosis is prospected.
【学位授予单位】：沈阳建筑大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TQ133.1
[Abstract]:Aluminum electrolysis process is one of many complex industrial processes. It has the characteristics of nonlinear, time-varying and large time-delay of complex industrial processes. In the process of aluminum electrolysis, there is not only the failure of electrolytic cell caused by the destruction of material balance and energy balance in the electrolytic cell, but also the fault caused by the fault of actuator motor in the process of electrolysis. In this paper, the fault of electrolysis cell is defined as the process fault of aluminum electrolysis, and the fault of actuator motor is defined as the fault of system. In this paper, a multi-fault diagnosis system for aluminum electrolysis is put forward, and the monitoring system of aluminum electrolysis is designed on the platform of VB 6.0. The multi-fault diagnosis system includes aluminum electrolysis process fault diagnosis system and aluminum electrolysis system fault diagnosis system. Aluminum electrolysis monitoring system includes electrolytic cell monitoring system and motor monitoring system. According to the above, I have done the following work. First, the significance of aluminum electrolysis fault diagnosis is expounded, the development status and existing problems of aluminum electrolysis fault diagnosis at home and abroad are summarized, and the research purpose and significance of this subject are described in allusion to the existing problems in aluminum electrolysis fault diagnosis. The author puts forward his own research method. Secondly, the working principle of aluminum electrolysis and the production process of aluminum electrolysis are summarized. Several common process faults in aluminum electrolysis process and common system faults in aluminum electrolysis process are discussed, and these faults are briefly introduced. Thirdly, the fault diagnosis model of aluminum electrolysis process is established, the basic principle of neural network is expounded, and the possibility and necessity of application of neural network in fault diagnosis are analyzed. The selection of characteristic quantity and the processing of input data in aluminum electrolysis process are discussed. The training process of neural network is optimized by using leapfrog algorithm. The feasibility of fault diagnosis model of aluminum electrolysis process is verified by simulation experiment. Fourthly, the fault diagnosis model of aluminum electrolysis system is established, and the superiority of using EMD algorithm to deal with the fault signal of motor is expounded. The wavelet analysis theory and wavelet packet analysis are used to extract the feature of fault signal. Combined with wavelet analysis and neural network, the feasibility of fault diagnosis model of aluminum electrolysis system is verified by simulation experiment. Fifthly, the monitoring software of aluminum electrolysis is designed based on VB 6.0. The general idea and design scheme of the software are summarized, and the realization process of the functions of each part of the monitoring software is introduced in detail. Sixth, the methods of aluminum electrolysis fault diagnosis used in this paper are summarized, and the future development of aluminum electrolysis fault diagnosis is prospected.
【学位授予单位】：沈阳建筑大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TQ133.1
【相似文献】
相关期刊论文 前10条
1 李R，
本文编号：2160889
本文链接：https://www.wllwen.com/kejilunwen/huaxuehuagong/2160889.html
