火电厂论文：火电厂锅炉智能故障诊断系统研究.docVIP

火电厂论文：火电厂锅炉智能故障诊断系统研究 【中文摘要】”四管”破裂是火电厂锅炉主要运行故障,影响着锅炉设备的安全运行,如何及时发现和防止“四管”破裂,智能故障诊断在火电厂运行中显得非常重要。本文主要研究了智能故障诊断的两种应用,它们分别是多维BP神经网络在锅炉故障诊断中的应用和结合数据挖掘与BP神经网络在锅炉故障诊断应用。其中多维BP神经网络在锅炉故障诊断中的应用是针对锅炉热工生产过程特点,采用多维BP神经网络方法进行锅炉故障诊断建模。BP神经网络采用多维结构,输入层采用模糊数学方法对运行参数进行量化,通过参数之间和参数与故障类之间的关联关系,建立多维BP神经网络模型,以锅炉“四管”泄漏为例,进行故障仿真实验,实验结果表明此方法能有效、快速和正确的诊断出结果;而结合数据挖掘与BP神经网络在锅炉故障诊断应用吸取两种方法的优点,把关联规则数据挖掘与BP神经网络方法有机的集成,建立一个智能故障诊断系统,更加准确的模拟人脑思维过程。也以火电厂锅炉“四管”破裂故障诊断为例,用数据挖掘方法从大量模糊和随机的历史数据中提取故障征兆与类型之间的关联关系数据,使数据有效的输入神经网络进行故障仿真实验。实验结果表明此方法也能有效、快速和正确的诊断出结果。两种方法各有千秋,在应用时要根据锅炉设备的特点对对象参数进行建模,选择最优诊断应用对故障进行预测诊断。 【英文摘要】”Four pipe” fracture is the main problem in the boiler operation which affect the safe operation of the boiler equipments Intelligent fault diagnosis plays an important role in power plant operation as to how to timely find and prevent the攆our多” burst.. This paper introduced two applications of intelligent fault diagnosis in the boiler, which are multi-dimensional BP neural network in fault diagnosis of the boiler and the combination of data mining and BP Neural Network Fault Diagnosis. The applications of BP neural network is the way of fault diagnosis modeling targeted at the feature of thermal characteristics of the boiler with the adoption of multi-dimensional method of BP neural network. The input layer used the fuzzy mathematic method to quantify the operating parameters and then established the Multi-dimensional structure of BP neural network through the relations between parameters and patterns of the fault. Taking the “Four pipe” fracture for instance, the BP neural network model was built and the results of the simulation experiment proved that this method can give a rapid, efficient result with high accuracy. Combining the advantage of data mining and BP Neural Network Fault Diagnosis in Boiler, integrating the rule mining and BP neural Network of organically, the intelligent fault diagnosis system was established which can demonstrate a more exact process of the human thought, and