摘要
建立精确的数学模型对于智能发电运行控制系统的设计与优化有着至关重要的作用,天牛须搜索(BAS)算法无须预知函数的梯度信息即可进行寻优,但该算法易早熟,且寻优精度不高,因此将混沌扰动机制与自适应因子引入该算法,通过对某1000MW火电机组零初始值与小波包数据处理后,结合算法对双输入单输出的电站锅炉尾部烟气含氧量模型进行系统辨识。针对实际电站测试数据,辨识结果表明:改进后的天牛须搜索算法在辨识时间上减少了43.00%,辨识误差上减小47.33%,辨识得到的传递函数在结构上符合理论分析,在参数上与机组实际运行工况相符,具有较好的机组分析研究价值。
The establishment of accurate mathematical model plays a crucial role in the design and optimization of intelligent power generation operation control system.Beetle Antennae Search(BAS)algorithm can perform optimization without knowing the gradient information of function in advance,however,due to the prematurity of the algorithm and the low precision of optimization,the chaotic disturbance mechanism and the adaptive factor are introduced into the Beetle Antennae Search algorithm.After processing the zero initial value and wavelet packet data of a 1000MW thermal power unit and combined with the algorithm,the oxygen content model of flue gas at the rear past end of power plant boiler with double input and single output was identified systematically.According to the actual test data,the identification results show that the improved Beetle Antennae Search algorithm is 43.00%less in identification time and 47.33%less in identification error than the previous algorithm.The transmission function obtained by identification conforms to theoretical analysis in structure and its parameters are consistent with the actual operating conditions of the unit,so it has good unit analysis and research value.
作者
孙宇贞
李帅
唐毅伟
SUN Yuzhen;LI Shuai;TANG Yiwei(College of Automation Engineering,Shanghai University of Electric Power,Shanghai 200090,China;Shanghai Engineering Research Center of Intelligent Management and Control for Power Process,Shanghai 200090,China)
出处
《锅炉技术》
北大核心
2022年第5期1-8,共8页
Boiler Technology
基金
上海市重点实验室建设项目(13DZ2273800)
上海市科学技术委员会工程技术研究中心项目(14DZ2251100)。
关键词
智能发电
天牛须搜索算法
混沌扰动机制
自适应因子
小波包数据处理
系统辨识
intelligent power generation
Beetle Antennae Search algorithm
chaotic disturbance mechanism
adaptive factor
wavelet packet data processing
system identification
