基于扩张状态卡尔曼滤波器的光伏/光热复合热泵系统经济模型预测控制研究  

Extended state Kalman filter-based economic model predictive control of solar photovoltaic/thermal composite heat pump system

作　　者：孙立 吴梦丹 郭萌萌 陈森 许文盈 虞文武 SUN Li;WU MengDan;GUO MengMeng;CHEN Sen;XU WenYin;YU WenWu(National Engineering Research Center of Power Generation Control and Safety(Southeast University),Nanjing 210096,China;School of Mathematics and Statistics,Shaanxi Normal University,Xi’an 710119,China;Jiangsu Provincial Key Laboratory of Networked Collective Intelligence,School of Mathematics,Southeast University,Nanjing 210096,China)

机构地区：[1]大型发电装备安全运行与智能测控国家工程研究中心(东南大学),南京210096 [2]陕西师范大学数学与统计学院,西安710119 [3]东南大学数学学院,江苏省网络群体智能重点实验室,南京210096

出　　处：《中国科学:技术科学》2025年第2期245-260,共16页Scientia Sinica(Technologica)

基　　金：国家自然科学基金项目(批准号:51936003,U22B6001);江苏省应用数学科学研究中心项目(编号:BK20233002)资助。

摘　　要：太阳能光伏/光热复合热泵(PV/T-HP)系统在绿色建筑及供热绿电替代方面具有广泛应用前景.然而,目前PV/T-HP系统设计往往基于稳态额定工况,未考虑运行过程中面临的源荷多源扰动、多变量耦合及动态能效衰减等问题.为此,本文提出了基于扩张状态卡尔曼滤波器(ESKF)的经济模型预测控制(EMPC)策略.首先,建立PV/T-HP系统的非线性机理模型,并基于部分工况下的线性化状态空间模型设计ESKF,实现系统状态与总扰动的联合实时估计.其次,基于ESKF扰动补偿,以系统能耗为优化目标设计了EMPC策略,并设定运行安全约束,实现控制量的滚动优化求解,证明了控制策略的可行性和最优性.最后,基于PV/T-HP系统实验平台验证了模型准确性,并对比了MPC和EMPC控制器的变负荷控制响应性能.实验测试结果显示,相比传统MPC,ESKF-EMPC方法可使系统耗电量降低8.1%,能效提升7.83%,验证了本文方案的有效性、鲁棒性与经济性.The solar photovoltaic/thermal composite heat pump(PV/T-HP)system has broad application prospects in green buildings and the replacement of heating and green electricity.However,the current design of PV/T-HP systems is often based on steady-state rated operating conditions,without considering issues such as multi-source disturbances,multivariable coupling,and dynamic energy efficiency degradation during operation.In view of this,this paper proposes an economic model predictive control(EMPC)strategy based on the extended state Kalman filter(ESKF).Firstly,we established a nonlinear mechanism model for the PV/T-HP system and designed an ESKF based on a linearized state space model under certain operating conditions to achieve a combined real-time estimation of the system state and total disturbance.Secondly,based on ESKF disturbance compensation and with system energy consumption as the optimization objective,an EMPC strategy was designed.Simultaneously setting operational safety constraints to achieve rolling optimization of control variables proves the feasibility and optimality of the control strategy.The experimental test results show that compared to traditional MPC,the ESKF-EMPC method can reduce system power consumption by 8.1%and improve energy efficiency by 7.83%,verifying the effectiveness,robustness,and economy of the proposed scheme.

关 键 词：光伏/光热复合热泵系统 经济模型预测控制 扩张状态卡尔曼滤波器 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]