摘要
针对孤网模式下水电机组PID控制策略鲁棒性差的缺点,提出基于双模型、实时反馈线性化、非线性特性补偿器和广义预测控制器的自适应广义预测控制策略(Adaptive Generalized Predictive Control,AGPC),以实现对水电站全工况实时最优控制。首先基于BP神经网络和改进樽海鞘算法构建高精度非线性水轮机模型,并结合PID控制器、引水系统、随动系统、发电机等模块搭建非线性水轮机调节系统仿真平台,然后通过定量计算揭示广义预测控制器直接应用于非线性水轮机调节系统的不合理性。进一步,结合实时反馈线性化和基于预估控制信号设计的水轮机非线性特性补偿器,提出孤网模式下广义预测控制器在水电站应用策略。最后,搭建了基于AGPC的非线性水轮机调节系统仿真平台,并以某电站真实数据验证了所提控制策略适用性、可靠性以及优良的鲁棒性。
The PID control strategy for hydropower units under frequency control mode has the disadvantage of poor robustness.To address this issue,this paper proposes an adaptive generalized predictive(AGPC)based on double models,real-time feedback linearization,nonlinear characteristic compensator,and generalized predictive controller to achieve real-time optimal control of hydropower stations under all operating conditions.Firstly,a high-precision nonlinear hydro-turbine model is constructed based on the BP neural network and the improved salp swarm algorithm,and a nonlinear hydraulic turbine regulation system simulation platform is built by combining modules such as the PID controller,diversion system,servo system,and generator.Then,through quantitative calculations,the irrationality of directly applying generalized predictive controller to the nonlinear hydraulic turbine regulation system is revealed.Furthermore,combining real-time feedback linearization with a nonlinear characteristic compensator designed based on predictive control signals,the application strategy of generalized predictive controller under frequency control mode is proposed for hydropower stations.Finally,a nonlinear hydraulic turbine regulation system simulation platform based on AGPC is built,and the applicability,reliability,and excellent robustness of the proposed control strategy are verified by the real data of a hydropower station.
作者
陈金保
刘少华
陈上
邹屹东
郑阳
肖志怀
CHEN Jinbao;LIU Shaohua;CHEN Shang;ZOU Yidong;ZHENG Yang;XIAO Zhihuai(School of Power and Mechanical Engineering,Wuhan University,Wuhan 430072,China;Key Laboratory of Hydraulic Machinery Transients,Ministry of Education,Wuhan University,Wuhan 430072,China;Nanjing Nanrui Jibao Electric Co.,Ltd,Nanjing 210000,China;China Yangtze Power Co.,Ltd.Three Gorges Power Plant,Yichang 443000,China)
出处
《水利学报》
EI
CSCD
北大核心
2024年第5期619-630,共12页
Journal of Hydraulic Engineering
基金
国家自然科学基金项目(51979204,51379160)
湖北省自然科学基金项目(2022CFD165)
中央高校基本科研业务费专项(2042022kf1022)。
关键词
广义预测控制
水电机组
反馈线性化
双模型
非线性特性补偿器
generalized predictive control
hydropower units
feedback linearization
double models
nonlinear characteristic compensator
