递阶稳态优化下非线性大工业过程的迭代学习控制

The Iterative Learning Control for Dynamics in Steady-state Hierarchical Optimization of Nonlinear Large-scale Industrial Processes

对递阶稳态优化下非线性大工业过程施行迭代学习控制 ,目的是进一步改善大工业过程的动态品质 .建立迭代学习控制的基本结构 ,提出迭代学习控制算法关于控制系统的ε-收敛性和期望目标轨线的δ -可达性的概念 ,对具有死区与滞后的饱和非线性大工业过程控制系统给出加权超前开环PD-型迭代学习算法 .利用 Bellman-Gronwall不等式和λ范数理论 ,论证了算法的收敛性 .数字仿真表明 ,迭代学习控制能有效改善递阶稳态下非线性大工业控制系统的动态品质 .

In this paper, the iterative learning control is studied for dynamics in steady-state hierarchical optimization of nonlinear large-scale industrial processes. The basic iterative learning control structure is established. The \$ε\$-convergence of the algorithms with respect to the control systems is defined, the \$δ\$-reachability of the desired trajectories is given. The weighted leading open-loop PD-type iterative learning control algorithm is suggested for saturated nonlinear large-scale industrial processes with dead zone and time delay. By means of Bellman-Gronwall inequality and \$λ\$-norm theory, the convergence of the algorithms is also proved. The numerical simulation shows that the iterative learning control can remarkably improve the dynamic performance of nonlinear large-scale industrial control system in steady-state optimization.