基于 PLC 的机舱报警监控系统的设计与故障诊断分析计算机专业.docx

基于 PLC 的机舱报警监控系统的设计与故障诊断分析 Design and Fault Diagnosis Analysis of Engine Room Alarm Monitoring System Based on PLC 摘要 机舱报警监控系统作为船舶自动化的重要部分，它能及时准确的反映出机舱内各动力设备、控制设备以及机电设备的运行状态，只要系统检测到所监测设备运行发生故障或者异常，该系统会触发报警机制，启动声光报警与并通过延伸报警板通知其他值班舱室，同时所有报警状态将被记录并打印出来。系统会实时监测机舱各个监控点的参数状态，各个监视屏上都会显示相关运行信息，使操作者能清楚的了解到船舶实时的运行状态。这些显示屏通常安放在集控室或者延伸报警面板上。本文以镇江亿华系统集成有限公司与天津航道局合作研发的“通途”号耙吸挖泥船改造项目为基础， 从理论与实践相结合的角度来设计出一套完整的机舱报警监控系统及其故障诊断系 统。 论文首先介绍了机舱报警监控系统的发展背景与意义，结合系统本身的基本功能与设计要求，阐述国内外对该系统的研究现状并给出总体的设计方案。系统下位机采用西门子的S7-400 作为控制器，其中数据采集与PLC 的CPU 模块之间采用PROFIBUS 总线通讯，上位机与 PLC 模块之间采用工业以太网实现数据共享。编程软件采用西门子公司开发的 Step-7 软件，实现硬件组态以及网络结构的创建，编程语言采用梯形图语言的方式，简单易懂，直接明了，整个系统采用模块化的编程思想，涉及到 OB 模块、DB 模块、FB 模块以及 FC 模块。 其次将从理论的角度设计出一套以神经网络算法为主的新型故障诊断系统。该部分将详细研究 BP 网络算法，RBF 网络算法以及 ELMAN 网络算法的模型结构、参数设置、训练函数以及每个函数各自的优缺点。其中故障诊断对象为船舶增压系统，通过对历史数据的拟合对网络模型进行训练，再将采集数据作为网络模型的输入，实现对系统故障的识别以及预测的效果。 最后将针对神经网络模型自身存在的缺点，设计出以遗传算法为主的优化方案， 利用遗传算法收敛性强的特点对神经网络的阈值以及权值进行优化，使得神经网络的预测精度提升，提高故障诊断的准确性，达到最优的预测结果。 关键词：监控系统 S7-400 总线通讯 神经网络 故障诊断 遗传算法 I Abstract The nacelle alarm monitoring system is an important part of ship automation. It can accurately and accurately reflect the operating status of various power equipment, control equipment and electrical equipment in the nacelle. As long as the system detects that the monitored equipment has malfunctioned or is abnormal, the system will trigger The alarm mechanism activates sound and light alarms and informs other duty cabins by extending the alarm board. All alarm conditions are recorded and printed out. The system will monitor the parameter status of various monitoring points in the cabin in real time. The relevant operation information will be displayed on each monitoring screen, so that the operator can clearly understand the real-time operating status of the ship. These displays are usually placed in a central control room or an extended alarm panel. Considering the backwardness of our country's relevant technologies in the current large environment, after reading the literature on the cabin monitoring system of the relevant ship, this paper will design a set of ship cabin alarm monitoring systems that are feasible and meet relevant regulations. The thesis first introduces the background and significance of the cabin alarm monitoring system, combines the basic functions and design requirements of the system itself, elaborates the research status of the system at home and abroad, and gives the overall design plan. The lower computer of the system adopts Siemens S7-400 as the controller, in which data acquisition and PLC CPU module use PROFIBUS communication between the host computer and the PLC module using industrial Ethernet to achieve data sharing. The programming software uses Step-7 software developed by Siemens to realize the hardware configuration and the creation of the network structure. The programming language adopts the ladder language method, which is easy to understand and straightforward. The whole system adopts a modular programming concept and involves OBs. Modules, DB modules, FB modules, and FC modules. After the system is designed, a new fault diagnosis system based on neural network algorithm will be designed from the theoretical point of view. This part will study the model structure, parameter setting, training function of BP network algorithm, RBF network algorithm and ELMAN network algorithm in detail as well as the respective advantages and disadvantages of each function. The fault diagnosis object is the ship supercharging system. The network model is trained by fitting the historical data, and the collected data is used as VII the input of the network model to realize the recognition of the system failure and the effect of the prediction. In the end, aiming at the shortcomings of the neural network model itself, an optimization scheme based on genetic algorithms is designed. The genetic algorithm is used to optimize the thresholds and weights of the neural network based on its strong convergence characteristics, so that the prediction accuracy of the neural network is improved. Improve the accuracy of fault diagnosis and achieve optimal forecast results. Keywords: Monitoring System, S7-400, Field bus Communication,NeuralNetwork,Fault Identification,Fault Prediction,Genetic Algorithm 目 录 摘要 I Abstract II 目 录 V Contents IX 第 1 章 绪论 1 1.1 课题研究背景及意义 1 1.2 机舱报警监控系统国内外发展现状 1 1.2.1 集中型监视系统 2 1.2.2 集散型监视系统 2 1.2.3 全分布式系统 3 1.2.4 现场总线型监控系统 3 1.3 船舶故障诊断系统国内外发展现状 3 1.3.1 智能化船舶诊断 4 1.3.2 信号处理技术 4 1.3.3 综合船舶诊断技术 4 1.4 课题主要研究内容 4 第 2 章 机舱报警监控系统总体结构设计 7 2.1 机舱报警监控系统的功能与要求 7 2.1.1 系统基本功能 7 2.1.2 技术标准以及规范要求 8 2.2 系统总体框架设计 9 2.3 机舱报警监控系统的硬件介绍 10 2.3.1 电缆的选型 10 2.3.2 UPS 10 2.3.3 网络设备 10 2.3.4 报警打印机 11 2.4 本章小结 10 第 3 章 机舱报警监控系统的下位机以及界面设计 13 3.1 下位机 PLC 介绍 13 3.1.1 PLC 概述 13 3.1.2 PLC 模块选择 14 3.2 PLC 程序以及控制流程设计 15 3.2.1 PLC 编程语言种类介绍 15 3.2.2 PLC 硬件组态 16 3.2.3 PLC 编程设计 18 3.2.4 控制流程设计 22 3.3 系统通讯网络结构设计 23 3.3.1 PROFIBUS 现场总线技术介绍