模式切换类经济预测控制切换时间在线估计

Online estimation of switching time for a class of optimization-control mode switch of economic model predictive control

针对一类非线性仿射系统,提出一种在线估计切换时间的经济模型预测控制算法,并将其拓展到长周期控制过程中。有限时间内,将切换时间作为变量实时更新估计,确定最优的切换操作点,以保证每一时刻都可以在控制目标可达的前提下经济性能最优,避免了传统切换经济预测控制策略可能出现的控制目标不可达或经济性能较差的情况。进一步,将该策略作为单周期应用到长周期优化控制过程中,当系统受到扰动时,开始一个新的优化控制周期,实现优化模式与控制模式的灵活切换,同时可以及时应对扰动的出现。该策略保证系统的综合性能最优,仿真结果证明了方法的有效性。

For a kind of nonlinear affine systems,an economic model predictive control algorithm for online estimation of switching time is proposed and extended to the long period control process.In a limited time,the switching time is used as a variable to estimate in real time,and the optimal switching point is determined,so as to ensure that the economic performance is optimal at every moment under the premise that the control objective is reachable.The situations of unreachable control objective and poor economic performance caused by traditional switching economic model predictive control strategy are avoided.Furthermore,the strategy is applied as a single cycle to the long period optimization control process.Once the disturbance occurs,a new optimization control cycle is started.The flexible switching between the optimized mode and the control mode can be realized,and the disturbance can be dealt with in time.The simulation results show the effectiveness of the method.