Design and application of multivariable digital robust optimal controllers
多变量数字鲁棒最优控制器的设计及应用

For linear multivariable discrete-time control systems, a new design approach to robust optimal controllers is developed in this paper. By presenting a new design concept to select row characteristic functions for a multi-variable system, a robust decomposition method is built up to guarantee the normalization of the closed-loop transfer function matrix. The H-infinity robust stability and the optimal time domain indices for the closed-loop system are then achieved simultaneously by utilizing the 1-infinity optimal control approach in its decoupling equivalent system. Finally, optimization and robustness of this digital controller are verified by an on-line application in a paper-making control system for three different indices of quantity and moisture.