约束Hammerstein系统输出反馈非线性预测控制

Output Feedback Nonlinear Predictive Control for Constrained Hammerstein Systems

针对具有状态、输入和中间变量约束的Hammerstein系统,采用两步法控制策略,给出一种新的可保证闭环系统指数稳定的输出反馈非线性模型预测控制算法。基于Ham-merstein系统的特殊结构,结合状态观测器给出无约束线性环节的输出反馈最优控制律,通过滚动优化一有限时域的约束优化问题计算实际控制量。给出保证闭环系统指数稳定的充分条件。以工业双环管聚丙烯装置牌号切换控制为例进行仿真,仿真结果验证该算法的有效性和实用性。

With the two-step control strategy, an output feedback nonlinear model predictive control （NMPC） algorithm to guarantee the exponential stability of closed-loop systems is developed for Hammerstein systems with constraints on states, inputs and intermediate variables. Based on the special structure of Hammerstein systems, an output feedback optimal control law is derived via state observation for the linear subpart without consideration of constraints. Actual control actions are computed in consideration of constraints by receding solution of a finite horizon optimal control problem. The sufficient condition for exponential stability of the closed-loop system is obtained. Finally, the simulation example of the grade transition control of industrial bi-tubular polypropylene plants shows the effective- ness and applications of the proposed algorithm.