基于灰色误差神经网络的热电厂燃炉系统故障预测应用研究

发布时间：2018-03-28 19:09

本文选题：燃炉系统　切入点：故障预测　出处：《青岛科技大学》2017年硕士论文

【摘要】：随着自动控制系统的普遍应用以及工业自动化规模的不断壮大,企业对系统运行的可靠性和稳定性提出了越来越高的要求。为了尽量减少甚至避免故障损失,提高各大型设备运行的安全性,对各种异常状态或故障状态做出及时准确的预测与诊断是非常有必要的。故障预测技术是随着维修理念的转变和维修方式的变革而逐步发展起来的新技术,其有效地实现了对故障的主动预防和实时监测,由基于智能系统的故障预测逐步替代了传统的事后诊断,一定程度上为实现设备的安全性和经济性运转提供了可行性。本文深入研究了热电厂锅炉的基本结构、工艺流程及主要安全性能指标,在此基础上建立了锅炉故障预测系统知识库。以4#炉高温过热器的壁温数值为研究样本,分别对T-S模糊神经网络结构模型和多参数灰色误差神经网络模型进行MATLAB仿真。实际验证结果表明了多变量灰色误差神经网络模型满足了较高的精度和速度的要求。为了深入研究该模型算法而设计了故障预测系统,简洁友好的界面和便捷的操作方式对于实现锅炉相关的故障预测具有一定的积极意义。某电厂的锅炉控制系统采用的是新华DCS的TisNet-OnXDC800,该系统本身集成了报警技术。本文深入研究了该套系统的运行机制与组态,为更好的实现对故障的识别与预测,运用OPC技术的通讯方式研究了预测模型和系统的结合。同时在XCU图形组态软件中研究了多参数灰色误差神经网络模型的合理运用,具有一定的实际意义。
[Abstract]:With the widespread application of automatic control system and the continuous expansion of industrial automation scale, enterprises put forward more and more high requirements for the reliability and stability of system operation. To improve the safety of the operation of large equipment, It is necessary to make timely and accurate prediction and diagnosis of various abnormal or fault states. Fault prediction technology is a new technology developed with the change of maintenance concept and maintenance mode. It effectively realizes the active prevention and real-time monitoring of the fault, and gradually replaces the traditional post diagnosis with the fault prediction based on the intelligent system. To some extent, it provides the feasibility for the safety and economical operation of the equipment. In this paper, the basic structure, process flow and main safety performance indexes of the boiler in the thermal power plant are deeply studied. On this basis, the knowledge base of boiler fault prediction system is established. The T-S fuzzy neural network model and the multi-parameter grey error neural network model are simulated by MATLAB. The results show that the multivariable grey error neural network model can meet the requirements of high precision and speed. In order to deeply study the model algorithm, a fault prediction system is designed, Simple and friendly interface and convenient operation mode have certain positive significance to realize boiler related fault prediction. The boiler control system of a power plant adopts TisNet-OnXDC800 of Xinhua DCS, which integrates alarm technology. In this paper, the operating mechanism and configuration of the system are studied. In order to better realize the fault identification and prediction, the combination of prediction model and system is studied by using the communication mode of OPC technology. At the same time, the rational application of multi-parameter grey error neural network model is studied in the XCU graphic configuration software. Has certain practical significance.
【学位授予单位】：青岛科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP183;TM621.2

【参考文献】