摘要
在压水堆核电站中,引起反应堆停堆的主要原因之一是蒸汽发生器(SG)的水位超过了安全界限。为将SG水位控制在一定范围内以保证核电站安全、可靠、经济地运行,将蒸汽流量作为可测扰动加入预测模型中,设计了具有前馈补偿作用的模型预测控制器(MPC)。对给水流量增量施加约束以减小对给水阀的冲击,采用二次规划(QP)最优化算法求解给水流量的最优值,求解时先对水位施加硬约束,若无可行解则加入软约束,以保证系统的稳定性。仿真结果表明,与变增益PID控制相比,MPC的控制效果更好。
Unexpected reactor shutdowns are often caused by violation of safety limits on the steam generator water level. A model predictive controller with feedforward compensation is designed to keep the water level in a certain range and enhance plant availability factor. The steam flowrate is included in the predictive model as measurable disturbance, and the constraint over feed-water flowrate increment is added. The quadratic program optimization algorithm with soft constraint is adopted to find a feasible quasi-optimal solution to guarantee the stability of the system. The simtdation results show that the model predictive control has better performances than the variable gain PID control.
出处
《控制工程》
CSCD
2008年第3期250-252,272,共4页
Control Engineering of China
基金
国家自然科学基金资助项目(60574038)
关键词
蒸汽发生器
水位控制
模型预测控制
前馈补偿
软约束
steam generator
water level control
MPC
feedforward compensation
soft constraint