基于数据驱动的浆液循环泵运行优化研究

浆液循环泵是火电厂石灰石-石膏湿法烟气脱硫系统中主要的耗电设备，实现浆液循环泵运行优化对降低电厂电耗、提高电厂经济效益具有重要的现实意义。针对当前火电厂浆液循环泵运行现状提出一种基于历史运行数据、采用聚类和分类相结合的浆液循环泵运行优化方法。采用模糊C均值聚类算法和基于层次分析法与熵权法加权融合的组合评价方法，对不同工况下浆液循环泵历史运行操作的自动寻优；以最优历史操作记录训练SVM分类器，使用遗传算法进行参数寻优，以实时运行数据为输入给出浆液循环泵的运行操作建议，从而实现其运行优化。以某电厂实测数据进行的仿真实验表明，使用该浆液循环泵运行优化方法，浆液循环泵耗电量最大可降低21.55%，对火电厂节能减排工作具有参考意义。

Research on Data-Driven Optimal Operation of Slurry Circulation Pump

As the main power consumption equipment in the limestone-gypsum wet flue gas desulfurization system of thermal power plants,optimizing the operation of the slurry circulation pump is of great practical significance to reduce power consumption and improve economic efficiency of the power plant.In this paper,according to the current operation status of slurry circulation pumps in thermal power plants,an operation optimization strategy of slurry circulation pumps based on historical operating data is proposed incorporated with the combination of clustering and classification method.Specifically,the fuzzy C-means clustering algorithm and the combinatory evaluation method based on the weighted fusion of analytic hierarchy process and entropy weight method are applied to automatically search the optimal historical operation of the slurry circulation pump under different operating conditions,such that the SVM classifier is trained with the best historical operation record.Then by using the genetic algorithm for parameter optimization and the real-time operation data as input,to implement the recommended operations of the slurry circulation pump,the optimization of its operation can be achieved.The simulation experiments based on the onsite measurement data from a power plant show that by using this slurry circulation pump operation optimization method,the power consumption of the slurry circulation pump can be reduced by about 21.55%,which can provide the kind of valuable reference for energy saving and emission reduction in thermal power plants.