On-Line Performance Monitoring And Optimization Of Cold-End System In Steam Turbine System

The research on on-line performance monitoring and optimization of cold-end system in steam turbine system plays a crucial role in mastering operation characteristics of cold-end system in steam turbine system and improving units'operation efficiency. Therefore, taking some 300MW unit cold-end system as the object, this paper established performance evaluation model of cold-end equipment and carried out the optimal operation of circulating water pump based on the realistic condition of unit. Meanwhile, optimization methods based on mechanism model and data mining were combined in this paper for the respective advantages.First of all, cold-end system performance evaluation model was established based on vacuum characteristic of condenser and thermal characteristic of cooling tower which is the foundation of model modification and circulating water pump optimization in post chapters.Secondly, a soft sensing model was established for the static head using data-driven model construction method in order to provide data support for the water flow rate modification of open circulating water system. A mixed grey-box model was established for improving precision based on the combination of least square support vector machine (LSSVM) regression and mechanism model. An objective function of economic index was set up for the circulating water pump optimal operation among fully operating regions.Furthermore, another circulating water pump optimal operation suggestion was provided by association rule mining based on data-mining theory. In order to avoid the impracticability of data mining model among fully operating regions, an intermediate variable integration prediction model was built based on the combination of mechanism model which have satisfactory generalization to achieve theoretic optimization and the data mining model which takes more field situation into account, along with the optimization effect analysis.Finally, the software system of on-line performance monitoring and optimization of cold-end system in steam turbine system was developed based on Visual Studio.Net platform and eDNA database.