
浏览全部资源
扫码关注微信
1. 杭州电子科技大学电子信息学院,浙江省,杭州市,310018
2. 浙江创维自动化工程有限公司,浙江省,杭州市,310012
3. 杭州电子科技大学自动化学院,浙江省,杭州市,310018
4. 浙江省电力公司,浙江省,杭州市,310008
5. 浙江省经济与信息化委员会,浙江省,杭州市,310008
纸质出版日期:2012
移动端阅览
刘登涛, 章坚民, 马国梁, 等. 基于电厂能效监测平台的机组特性在线识别与可视化[J]. 电力系统自动化, 2012,36(8):66-70.
基于省级热电厂能效监测平台的运行数据,采用M5′模型树方法在线识别和绘制了较为精确、详细的单抽和双抽机组工况图。该方法采用分段式线性拟合,自动优选分段数并逼近可能非凸或非连续的非线性耗量曲线。利用该方法进行机组定期滚动识别对比,可以更好地察觉机组特性的变化,为机组能效监视提供可视化感知工具,为测量数据提供校验依据,并为电厂热电负荷优化分配和电网节能调度提供精确的分段线性化机组特性函数。
The M5′ model tree algorithm is used to on-line identify and draw the accurate and detailed single and double extraction unit working condition charts. This method automatically and optimally selects the number of sections and approaches the nonlinear consumption curves might be with non-convex and discontinuous property using piecewise linearization fitting. By comparison of scrolling identified characteristics at a fixed period, the change of unit can be detected and an awareness visualization tool of unit can be achieved. This method can provide the check basis for measured data and provide accurate piecewise linearization function of unit characteristics for load optimal sharing of thermal power plant and energy saving scheduling of the power grid.
0
浏览量
3087
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621