基于流体网络方法的电站锅炉热力系统建模与性能分析预测研究

发布时间：2018-02-11 14:08

本文关键词： 流体网络拓扑结构关联矩阵超临界锅炉热力计算回转式空预器软测量神经网络性能预测软件架构蒸汽温度吹灰优化　出处：《浙江大学》2014年博士论文　论文类型：学位论文

【摘要】：本文结合超临界电站锅炉性能分析平台,采用流体网络方法建立锅炉热力系统的一般数学模型,包括基于关联矩阵描述的锅炉热力系统拓扑结构和锅炉换热部件算法模型,并提出通用的模型求解算法。同时,提出了一种能够解决带有冷段和热段的三分仓回转式空气预热器传热计算方法,建立了传热模型并设计计算机程序和算法进行实现。针对部分电站锅炉热工参数不易测量的情况,采用软测量方法对其进行间接测量,并结合锅炉热力系统机理模型,建立电站锅炉的性能分析和在线预测系统,实现对锅炉机组的性能预测和控制。在系统实现技术上,采用基于浏览器／服务器的架构模式,研发支持在线监测的电站锅炉性能分析与预测系统,并应用于锅炉蒸汽温度的预测和吹灰优化之中。论文主要研究内容如下：介绍了流体网络的支路和节点的概念,给出了用图论中的关联矩阵、邻接矩阵和回路矩阵表示流体网络的方法。介绍了锅炉的工作原理和组成结构,分析了锅炉的热力系统,并用关联矩阵和邻接矩阵表示出了锅炉热力系统中烟气流程、空气流程和水流程的网络拓扑结构。介绍了锅炉热力计算的理论基础,给出了烟气流程、空气流程、水流程之间和各流程内部所满足的质量平衡方程、能量平衡方程、动量平衡方程,以及传热方程和热量平衡方程。在此基础上借用电网理论的基尔霍夫第一定律和基尔霍夫第二定律,建立起锅炉热力系统的数学模型,并给出了模型的求解方法。之后研究了三分仓回转式空气预热器热力计算的解析方法。火电厂许多重要技术参数和经济参数难以进行在线测量,选取与之有关的且便于测量的参数作为辅助变量,基于锅炉运行状态监测系统、机组运行历史数据库和煤质在线监测等系统,采用机理方法或历史数据建立软测量模型,从而实现对不易直接测量的锅炉热工参数进行软测量。在建立软测量模型上,采用BP神经网络建立基于历史数据的软测量模型对飞灰含碳量进行预测；对于受热面的灰污热阻,则主要采用机理分析的方法建立软测量模型。将软测量结果再次作为锅炉热力系统机理模型的输入,从而建立锅炉性能分析计算和预测模型。论述基于流体网络方法的锅炉热力系统的实现。先确定了锅炉的整体技术方案和系统架构,介绍了系统的主要功能和关键技术。对超临界锅炉机组性能分析系统进行了应用：利用所建立的锅炉性能分析与预测系统,对一台超临界电站锅炉的性能进行预测计算,并根据预测值,实现蒸汽温度的喷水减温控制。提供了锅炉积灰和结渣的监测和诊断方法,分析了锅炉积灰和吹灰的经济性,给出了最佳的吹灰周期,并采用模糊综合评判方法对锅炉的吹灰判定和吹灰顺序进行了优化。
[Abstract]:This paper combined with the supercritical power plant boiler performance analysis platform, the general mathematical model of boiler thermodynamic system with the fluid network method, including based on correlation matrix description of the boiler thermal system topology and boiler heat exchange component model, and put forward the general model algorithm. At the same time, put forward a kind of method to calculate the settlement with cold and the hot section of the three sectional regenerative air preheater, establishes a heat transfer model and computer program design and algorithm implementation. Aiming at the boiler thermal parameters is not easy to measure, the soft measurement method for indirect measurement of the model, and combined with the boiler thermal system mechanism, performance analysis and prediction system of power plant boiler online the performance, prediction and control of the boiler. In the realization of technology system, using the browser / server architecture based on Model The system is used to predict the steam temperature of the boiler and optimize the soot blower, which supports the on-line monitoring of the performance analysis and prediction of the boiler performance.
The main contents of this paper are as follows:
Introduces the concept of nodes and branches of fluid network, gives the correlation matrix in graph theory, adjacency matrix and loop matrix representation method of fluid network. This paper introduces the working principle and structure of the boiler, the boiler thermal system analysis, and using the correlation matrix and the adjacency matrix of the flue gas flow of boiler thermal system the air flow and water flow, the network topology structure.
Introduces the basic theory of the thermal calculation of boiler, flue gas flow, air flow and water flow are given, and the internal flow between the mass balance equation, energy balance equation, momentum balance equation and heat transfer equation and heat balance equation. Based on the theory of Kirchhoff's first law to borrow power and Kirchhoff's second law, mathematical model establish the boiler thermal system, and gives the method of solving the model. After the study of analytical method of three sectional regenerative air preheater thermal calculation.
Thermal power plant, many important technical and economic parameters are difficult to measure online, select related and easily measured parameters as auxiliary variables, the boiler operation condition monitoring system based on unit operation history database and coal quality online monitoring system, using the soft measurement modeling method and mechanism of historical data, so as to realize the soft measurement of is not easy to measure directly. The boiler thermal parameters in a soft measurement model, the soft measurement model of historical data to predict the carbon content of fly ash based on BP neural network; the heating surface fouling resistance, mainly by the mechanism analysis method to establish the soft measurement model. The soft measurement result again as a boiler the mechanism model of input, calculation and analysis so as to establish the prediction model of boiler performance.
The realization of the boiler thermal system based on fluid network method. To determine the boiler overall technical scheme and system architecture, introduces the main functions and key technologies of the system. The application of Supercritical Boiler Performance Analysis System: using the boiler performance analysis and forecast system, predict performance of a supercritical power station boiler, and according to the forecast, the water temperature of the steam temperature reducing control is provided. The monitoring and diagnosis methods of boiler ash and slag, analyzed the boiler ash and ash blowing, blowing cycle gives the best, and the fuzzy comprehensive evaluation method for boiler blowing determination of ash and soot blowing sequence is optimized.

【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM621.2

【参考文献】