基于贝叶斯网络的水泥回转窑故障诊断算法的研究

发布时间：2018-04-03 01:32

本文选题：贝叶斯网络　切入点：参数学习　出处：《燕山大学》2015年硕士论文

【摘要】：水泥回转窑是新型干法水泥生产线上的核心设备之一,其运行情况的好坏直接影响着水泥的产量和质量,能源消耗以及环境污染。目前,回转窑的故障诊断主要是依靠人工进行,难以高效准确的完成回转窑故障诊断。因此,本文在分析回转窑的煅烧工艺和主要工艺参数的基础上,开展了水泥回转窑故障诊断算法的研究与贝叶斯网络的水泥回转窑故障诊断网络和故障预测网络的构建,具体研究工作如下:首先,针对在不完整数据条件下贝叶斯网络参数学习准确率较低的问题,提出了一种基于改进遗传算法的贝叶斯网络参数学习方法—GSA算法。其次,通过分析专家知识和回转窑的煅烧工艺确定水泥回转窑贝叶斯故障诊断网络中的变量。由于回转窑中数据难以收集,且经常出现数据的缺失,故利用SEM算法在不完整数据条件下对网络进行结构学习,采用GSA算法在不完整数据条件下对网络进行参数学习,然后应用训练后的故障诊断网络对回转窑的故障进行诊断分析。最后,对基于贝叶斯网络的水泥回转窑的故障预测进行了研究和探讨,分析了利用贝叶斯网络进行水泥回转窑故障预测的基本流程,提出了一种确定根节点先验概率分布的方法。对窑内结圈故障的预测模型进行构建,并利用预测模型对测试数据集进行了预测。
[Abstract]:Cement rotary kiln is one of the core equipments in the new dry cement production line. Its operation directly affects the output and quality of cement, energy consumption and environmental pollution.At present, the fault diagnosis of rotary kiln mainly depends on manual, it is difficult to complete the fault diagnosis of rotary kiln efficiently and accurately.Therefore, on the basis of analyzing the calcination technology and main technological parameters of rotary kiln, the paper studies the fault diagnosis algorithm of cement rotary kiln and constructs the fault diagnosis network and fault prediction network of cement rotary kiln based on Bayesian network.The specific research work is as follows: firstly, a Bayesian network parameter learning algorithm GSA based on improved genetic algorithm is proposed to solve the problem of low accuracy of Bayesian network parameter learning under incomplete data.Secondly, by analyzing the expert knowledge and the calcination process of rotary kiln, the variables in Bayesian fault diagnosis network of cement rotary kiln are determined.Because the data in rotary kiln is difficult to collect and the data is often missing, the SEM algorithm is used to study the network structure under the condition of incomplete data, and the GSA algorithm is used to study the parameters of the network under the condition of incomplete data.Then the trained fault diagnosis network is used to diagnose and analyze the fault of rotary kiln.The prediction model of ring junction fault in kiln is constructed, and the test data set is predicted by the prediction model.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TQ172.622

【参考文献】