摘要
以大数据分析为基础,针对实际运行数据,研究非均匀工况划分方法,取得不同工况的典型数据;改进传统滑压曲线只是负荷的单值函数的缺点,应用卷积神经网络,建立负荷、主蒸汽温度、再热蒸汽温度、环境温度和主蒸汽压力的非线性模型,进而得到主蒸汽压力的实时优化值。经过验证,模型在精度和规律性上都取得了满意的效果。最后将主蒸汽压力优化模型应用到实际300 MW火电机组上,并进行闭环控制。结果证明:应用实时主蒸汽压力优化能够有效降低机组能耗,并且在不同环境温度下,有更高的节能潜力。
Based on the analysis of big data, this paper studies the method of dividing non-uniform working conditions and obtains the typical data of different working conditions according to the actual operating data. The traditional sliding pressure curve is only the single-valued function of load. The convolution neural network is used to establish the load, the main steam temperature, the reheat steam temperature, the ambient temperature and the main steam pressure nonlinear model, then the main steam pressure real-time optimization value is obtained. After verification, the model has achieved satisfactory results on accuracy and regularity. Finally, the main steam pressure optimization model is applied to the actual 300 MW thermal power unit, and the closed-loop control is carried out. The results show that the application of real-time main steam pressure optimization can effectively reduce unit energy consumption and has higher energy saving potential under different ambient temperatures.
作者
耿娜
夏志
王春玲
黄振群
冯忠宝
王松寒
金春林
GENG Na;XIA Zhi;WANG Chun-ling;HUANG Zhen-qun;FENG Zhong-bao;WANG Song-han;JIN Chun-lin(Electric Power Research Institute of State Grid Jilin Electric Power Co.Ltd.,Changchun 130021,China)
出处
《控制工程》
CSCD
北大核心
2020年第7期1223-1230,共8页
Control Engineering of China
关键词
主蒸汽压力
实时优化
大数据
卷积神经网络
闭环控制
Main steam pressure
real-time optimization
big data
convolution neural network
closed-loop control
