[关键词]
[摘要]
为了提高汽轮机能耗预警水平,将多变量状态估计技术(MSET)作为汽轮机能效监测的数据挖掘方法,在完成不同工况机组能耗水平分类的基础上,利用改进的熵权法计算能效偏离度。将偏离度和警报阈值结合来判断汽轮机运行工况水平,及时获取能耗时间点,并定位引起能耗异常的特征参数,该方法克服了单一热耗率指标评价能效变化的局限性。以某电站600 MW机组的汽轮机历史运行数据为研究对象,借助聚类算法完成汽轮机热耗分类,以最佳热耗率所在类簇的部分运行数据作为训练样本建立MSET能耗模型,并完成模型正确性验证。利用剩余类簇在能耗模型下实际观测值和模型估计值的偏差,并结合信息熵权法分配与热耗相关的特征参数属性权重,计算获得偏离度指数,完成汽轮机能耗异常预警。结果表明:该方法可以为汽轮机能效异常变化提供及时的预警信息并定位导致能耗异常的参数。
[Key word]
[Abstract]
In order to improve the early warning level of steam turbine energy consumption, multivariable state estimation technology (MSET) was taken as the data mining method for steam turbine energy efficiency monitoring. Based on the classification of unit energy consumption level under different working conditions, the improved entropy weight method was used to calculate the deviation degree of energy efficiency. The deviation degree and early warning threshold were combined to judge the operating condition level of steam turbine, obtain the time point of energy consumption in time, and locate the characteristic parameters causing abnormal energy consumption. This method overcame the limitation of evaluating the change of energy efficiency with a single heat rate index. The historical operation data of 600 MW steam turbine of a power station were then taken for the case study. Classification of steam turbine heat consumption was completed by using clustering algorithm. The MSET energy consumption model was established with part of the operation data in the best heat rate cluster as the training samples, and the correctness of the model was verified. The deviations were then obtained between the actual observation values and the estimated values of the model of the residual clusters under the energy consumption model. The attribute weight values of the characteristic parameters related to the heat consumption were allocated by the information entropy weight method. Moreover, the deviation degree values were calculated to complete the abnormal energy consumption early warning of steam turbine. The results show that this method can provide timely warning information for the abnormal energy efficiency change and position the parameters causing abnormal energy efficiency for steam turbine.
[中图分类号]
TK262
[基金项目]
国网江苏省电力有限公司科技项目(JF2021036)