基于故障树反演的JH39机械压力机故障诊断系统研究

发布时间：2018-03-22 16:05

本文选题：JH39-800机械压力机　切入点：故障树反演　出处：《合肥工业大学》2017年硕士论文　论文类型：学位论文

【摘要】：本文针对JH39-800机械压力机故障种类较多且难以诊断的特点,研究和讨论了一种故障诊断方法并建立故障诊断系统,能够对其润滑系统的异常情况迅速做出判断,从而及时发现故障。本文从JH39-800机械压力机各个部件的工作原理入手,熟悉整个系统,详细了解润滑系统功能及原理,收集基本信息和数据,分析出各部件的常见故障并且探索故障原因,建立故障树并进行定性分析和定量分析。构建BP神经网络算法,在MATLAB中对BP神经网络进行训练和仿真,根据检测到的故障信号,进行故障诊断,再运用故障树反演的方法计算故障概率。建立JH39-800机械压力机故障诊断的数据库系统,以数据库为基础,BP神经网络为核心,编写故障诊断程序,结合TWINCAT2自动化软件构建出基于BECKHOFF控制系统的故障诊断系统。通过BECKHOFF相关模块对故障信号进行处理,带入诊断系统中进行分析,得出诊断结果。将实验结果与仿真结果进行对比,分析诊断系统对各种故障信号的识别能力,判断该故障诊断系统的诊断效果。经过实验验证,该故障诊断系统对于故障信号具有一定的识别能力,能够有效地进行故障诊断,有助于在线判断JH39-800机械压力机内部的多种故障。同时,此方法也适用于其他同类型机械压力机故障诊断。
[Abstract]:This paper studies and discusses a fault diagnosis method and establishes a fault diagnosis system for JH39-800 mechanical press, which has many kinds of faults and is difficult to diagnose, which can quickly judge the abnormal condition of its lubrication system. This paper starts with the working principle of each component of JH39-800 mechanical press, is familiar with the whole system, understands the function and principle of lubrication system in detail, collects basic information and data. The common faults of each component are analyzed and the causes are explored, the fault tree is established, and the qualitative analysis and quantitative analysis are carried out. The BP neural network algorithm is constructed, and the BP neural network is trained and simulated in MATLAB. According to the detected fault signal, the BP neural network is trained and simulated. The fault diagnosis is carried out, and the fault probability is calculated by using the method of fault tree inversion. The database system of fault diagnosis for JH39-800 mechanical press is established, and the fault diagnosis program is compiled based on the database and BP neural network. The fault diagnosis system based on BECKHOFF control system is constructed by combining with TWINCAT2 automation software. The fault signal is processed by BECKHOFF related module, and the fault signal is brought into the diagnosis system for analysis, and the diagnosis result is obtained. The experimental results are compared with the simulation results. The ability of the fault diagnosis system to identify various fault signals is analyzed and the diagnosis effect of the fault diagnosis system is judged. The experimental results show that the fault diagnosis system has a certain ability to identify the fault signal and can effectively diagnose the fault. This method is helpful for on-line judgment of various internal faults of JH39-800 mechanical press, and it is also suitable for fault diagnosis of other mechanical presses of the same type.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG305;TP277

【相似文献】