基于数据挖掘的火电机组能耗特征提取

发布时间：2018-09-08 12:15

【摘要】：在煤炭资源供应日益紧张、煤炭价格上涨的严峻形势下,加强挖掘燃煤火力发电机组的能耗特性,是实现机组节能运行的重要基础,也是实现电力系统节能工作的一个重要方向,更是保障我国经济可持续发展的一个关键问题。因此,做好火电机组节能工作的基础性研究具有重要意义。针对此课题,论文主要工作如下：首先,利用因子分析法降维的功能挖掘出影响机组能耗的重要指标。在火力发电机组的历史运行数据的基础上,选取一组尽可能反映机组运行状态且与机组发电煤耗密切相关的参数。建立了机组能耗指标的公因子分析模型,并对所提取的公因子数目进行校验,同时考虑因子分析法应用中经常被忽略的特殊因子,通过特殊因子方差估计值大小去间接地判断所提取的机组能耗公因子分析模型是否出现Heywood现象,进一步对所拟合出的能耗公因子模型的优劣性进行定量判断。为了更有清晰地描述原始变量与少数公因子之间的关系,通过回归法计算因子得分,实现用少数公因子描述原始数据信息结构的目的,达到了降维的目的。其次,在火电机组目标值挖掘过程中,利用数据挖掘技术中的加权模糊C均值聚类方法对机组能耗的重要运行可控参数进行目标值的求解。考虑到机组不同能耗指标的样本点对分类的影响不尽相同,因而在机组参数挖掘的过程中不同属性的指标对聚类的贡献程度会有所差异,论文提出首先对机组的不同能耗指标赋予不同的权重,再进行目标值的挖掘,所挖掘出来的目标值更具合理性。最后,为了提高火电机组能耗评估的客观性和科学性,论文提出利用层次分析法与熵权法相结合来确定机组能耗评估中各项指标的权重值,克服了单一赋权法的缺点,使权重的确定更为合理。在此基础上,利用两级模糊综合评估方法对机组的能耗水平进行评估,同时对比主观赋权(层次分析法)得到权重的两级模糊综合评估结果。实例分析表明：该评估方法得出的评估结果更加完善和准确,更符合火电机组能耗水平的实际情况。论文通过火电机组的实时工况进行分析,对机组的能耗特征进行提取,得到的分析结果与理论知识是相符合的,验证了所提模型和问题分析思路的有效性和实用性,为实现火电机组的节能优化运行和调整奠定了一定的基础。
[Abstract]:Under the severe situation that the supply of coal resources is increasingly tight and the coal price is rising, it is an important foundation for realizing the energy-saving operation of the units to strengthen the excavation of the energy consumption characteristics of coal-fired generating units. It is also an important direction to realize the energy saving of power system and a key problem to ensure the sustainable development of our country's economy. Therefore, it is of great significance to do a good job of energy-saving work of thermal power units. The main work of this paper is as follows: firstly, the important indexes which affect the unit energy consumption are excavated by the function of reducing dimension by factor analysis method. On the basis of the historical operation data of thermal power generating units, a group of parameters which reflect the operating state of the units and which are closely related to the coal consumption of generating units are selected. The common factor analysis model of unit energy consumption index is established, and the number of common factors extracted is verified, and the special factors which are often neglected in the application of factor analysis are considered. The Heywood phenomenon is indirectly judged by the estimated value of the variance of special factors, and the advantages and disadvantages of the fitted common factor model of energy consumption are further quantitatively judged. In order to more clearly describe the relationship between the original variables and a few common factors, the purpose of describing the information structure of the raw data by using a few common factors is realized by calculating the factor scores by the regression method, and the purpose of reducing the dimension is achieved. Secondly, in the process of target value mining of thermal power units, the weighted fuzzy C-means clustering method in data mining technology is used to solve the target values of important operating controllable parameters of unit energy consumption. Considering that the sample points of different unit energy consumption indexes have different effects on classification, the contribution of different attribute indexes to clustering will be different in the process of unit parameter mining. The paper puts forward that the different energy consumption indexes of the unit are given different weights at first, and then the target value is excavated, and the target value is more reasonable. Finally, in order to improve the objectivity and scientific nature of energy consumption assessment of thermal power units, the paper proposes to use AHP and entropy weight method to determine the weight value of each index in unit energy consumption evaluation, which overcomes the shortcoming of single weighting method. Make the determination of weight more reasonable. On this basis, the two-stage fuzzy comprehensive evaluation method is used to evaluate the energy consumption level of the unit, and the weighted two-stage fuzzy comprehensive evaluation results are obtained by comparing the subjective weight (AHP). The analysis of examples shows that the evaluation results obtained by this method are more perfect and accurate, and more in line with the actual situation of energy consumption level of thermal power units. Through the analysis of the real time working conditions of thermal power units, the energy consumption characteristics of the units are extracted, and the results are in accordance with the theoretical knowledge. The validity and practicability of the proposed model and problem analysis ideas are verified. It lays a foundation for energy-saving optimization and adjustment of thermal power units.
【学位授予单位】：长沙理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM621;TP311.13

【共引文献】