基于历史数据挖掘的汽轮机阀门管理曲线优化

Optimization for Management Curves of Turbine Valves based on Historical Data Mining

汽轮机阀门管理曲线优化能够大大提高机组负荷控制精度和一次调频能力,但传统的试验方法较复杂、耗时较长,而常用的优化计算方式效率较低、准确性欠佳。运用数据挖掘方法,实现了历史运行数据驱动的汽轮机阀门流量特性计算,并通过仿真对DEH阀门管理曲线进行全行程的优化修正。优化后,流量特性的线性度和连续性有了很大改善,AGC变负荷和一次调频的能力得到了提高。

The optimization for the management curves of turbine valves can greatly improve the load control accuracy and primary frequency regulation performance of unit, but the traditional testing method is complicated and time-consuming, and the commonly used optimization methods have low efficiency and poor accuracy. The data mining method is used to calculate the flow characteristics of steam turbine valve driven by the historical operation data, and the whole stroke of DEH valve management curve is optimized by simulation. After the optimization, the linearity and continuity of the flow characteristics have been improved, and the capability of the load variation and primary frequency regulation has been enhanced.