首页|基于平方根容积卡尔曼滤波的核反应堆功率H∞控制

基于平方根容积卡尔曼滤波的核反应堆功率H∞控制

扫码查看
核反应堆的负荷跟踪控制是核电站灵活调峰、具有应用价值的基础之一。根据反应堆中子动力学、热工水力、反应性方程等数学模型构建了反应堆的平方根容积卡尔曼滤波器(Square-root Cubature Kalman Filter,SCKF),对反应堆状态反馈问题中部分状态量无法量测或估计精度不足的问题加以解决,结合离散H∞性能指标设计了基于SCKF的反应堆功率H∞控制器,并与基于经典卡尔曼滤波器(Kalman Filter,KF)的反应堆功率H∞控制器进行了同工况数值仿真对比。仿真结果表明,基于SCKF的功率控制器具有较好的负荷跟踪性能,能够根据期望跟踪目标调节反应堆功率水平;由于SCKF对非线性系统模型相比于KF具有更高的状态估计精度,基于SCKF的功率控制器相比基于KF的功率控制器具有更优的控制效果。
Discrete H∞ control of nuclear reactor power based on Square-root Cubature Kalman Filter
The nuclear reactor's load-following control was one of the foundations for nuclear power plants'flexible peak-adapting,and also one of the reasons of nuclear power plants's application value.In this paper,the Square-root Cubature Kalman Filter(SCKF)of the nuclear reactor,which had solved the problem that some of the nuclear reactor's state variables can not be measured or estimated accurately,was constructed based on mathematical models such as reactor neutron dynamics equation,thermo-hydraulics equation,and reactivity equation,and the reactor power H∞controller based on SCKF was designed by combining the discrete H∞ performance index,and compared with the reactor power H∞ controller based on the classical Kalman Filter(KF)through numerical simulationunder the same operating conditions simultaneously.The simulation results showed that the power controller based on SCKF had good load-following performance,and the controller based on SCKF was able to regulate the reactor power level according to the expected targets;due to the higher state estimation accuracy of SCKF compared with KF in nonlinear system models,the power controller based on SCKF had the better control performance compared with the controller based on KF.

nuclear reactorKalman FilterSquare-root Cubature Kalman FilterH∞ performance indexload-following control

张玉衡、王俊玲、刘雨昆

展开 >

核安全与仿真技术国防重点学科实验室 哈尔滨工程大学,哈尔滨 150001

核反应堆 卡尔曼滤波 平方根容积卡尔曼滤波 H∞性能指标 负荷跟踪控制

国家重点基础研究项目

5140504A0503

2024

哈尔滨商业大学学报(自然科学版)
哈尔滨商业大学

哈尔滨商业大学学报(自然科学版)

影响因子:0.405
ISSN:1672-0946
年,卷(期):2024.40(2)
  • 13