直接空冷机组凝汽器动态系统参数辨识
张珈铭¹，田静菲²，吴哲³，杨婷婷¹，邓慧⁴
（1. 华北电力大学控制与计算机工程学院，北京 102206；2. 天津国能津能滨海热电有限公司，天津 300459；3. 吉林电力股份有
限公司白城发电公司，吉林 白城 137000；4. 暨南大学能源电力研究中心，广东 珠海 519070）
摘要： 直接空冷机组凝汽器是受机组负荷与自然风扰动的复杂热力系统，准确预测直接空冷机组汽轮机背压的动态
响应特性，对于背压控制器的优化设计尤为重要。以一660 MW 直接空冷机组实际运行数据为参数样本，基于惯性权重
自适应的粒子群优化算法对空冷机组背压动态系统模型进行参数辨识，验证变工况动态模型的鲁棒性，进而设计前馈控
制器，以快速响应机组负荷对背压的扰动。结果表明：在不同工况下，基于惯性权重自适应调节的粒子群优化算法辨识
出的参数模型可以准确描述直接空冷机组的负荷、空冷风机转速指令与汽轮机背压之间的动态过程特性；前馈控制可有
效减小机组负荷对汽机背压的不利扰动，提高控制系统调节快速性。
关键词： 直接空冷机组；惯性权重自适应；粒子群算法；动态系统辨识；前馈控制器
中图分类号： TK264.11；TK39     文献标识码： A     文章编号： 2095-0802-(2021)11-0089-06
Parameter Identification of Condenser Dynamic System of Direct Air-cooled Unit
ZHANG Jiaming¹, TIAN Jingfei², WU Zhe³, YANG Tingting¹, DENG Hui⁴
(1. School of Control and Computer Engineering, North China Electric Power University, Beijing 102206, China; 2. Tianjin
Guoneng Jinneng Binhai Thermal Power Co., Ltd., Tianjin 300459, China; 3. Baicheng Power Generation Company, Jilin Electric
Power Co., Ltd., Baicheng 137000, Jilin, China; 4. Energy and Power Research Center, Jinan University, Zhuhai 519070,
Guangdong, China)
Abstract: The condenser of a direct air-cooled unit is a complex thermal system disturbed by unit load and natural wind.
Accurately predicting the dynamic response characteristics of the turbine back pressure of a direct air-cooled unit is particularly
important for the optimal design of the back pressure controller. Taking the actual operating data of a 660 MW direct air-cooled
unit as a parameter sample, based on the inertial weight adaptive particle swarm optimization algorithm, this paper parameterized
the air-cooled unit back pressure dynamic system model to verify the robustness of the variable-condition dynamic model, and
then design the feedforward controller to respond quickly to the disturbance of the unit load to the back pressure. The results show
that under different working conditions, the parameter model identified by the particle swarm optimization algorithm based on the
adaptive adjustment of inertial weight can accurately describe the dynamic process characteristics of the direct air-cooled unit
load, the air-cooled fan speed command and the steam turbine back pressure; feedforward control can effectively reduce the
unfavorable disturbance of unit load to turbine back pressure and improve the speed of control system adjustment.

Key words: direct air-cooled unit; inertial weight adaptive; particle swarm algorithm; dynamic system identification; feedforward
controller