摘要
为提高水电站参与电力市场竞争的能力,须建立水电站厂内经济运行准实时系统。结合不同类型水轮机组的特性,探讨了水电站厂内经济运行准实时系统任务时限的确定,并就水电站厂内经济运行准实时系统的一个主要难点问题———水轮机组耗水量的在线计算,提出一种在线训练、在线应用的神经网络计算模型。实例研究表明,采用神经网络方法实现水轮机组耗水量的在线计算,不仅能提高计算精度,而且能够满足水电站厂内经济运行的实时性要求。
In order to enhance the competition capability of hydropower plant in power market, the quasi real-time system of economical operation of hydropower plant (EOHP) should be built. According to the properties of different types of water turbine generator units, the deadline determination of the quasi real-time system task of EOHP is discussed. An online-training neural network model is proposed to calculate the flow consumption of water turbine generator units, which is the most difficult problem of the quasi real-time system of EOHP. The case study shows that the online calculation of the flow consumption by the online-training neural network model can improve the calculation precision and meet the requirement of the real-time system of EOHP.
出处
《中国农村水利水电》
北大核心
2005年第3期84-86,共3页
China Rural Water and Hydropower
基金
国家重点基础发展规划"973"项目(G1999043608)。
关键词
厂内经济运行
准实时系统
耗水量
负荷分配
任务时限
inner-plant economical operation
quasi real-time system
flow consumption
load distribution
deadline of task
