预测控制中静态目标的实现*

The Realizations of Steady State Object Inside Predictive Control

中文关键词  预测控制  静态目标  自由度  多目标优化  双重控制

英文关键词  predictive control  steady state object  degree of freedom  multiobjective optimization  dual control

中文摘要

      通过滚动优化性能指标进行非方系统处理，是预测控制主要优点之一．对“瘦系统”，可直接获得最小二乘解；而对“胖系统”，可能出现多余控制自由度，人们希望在满足基本控制要求的基础上，还能进一步实现一定的经济指标．由于经济指标一般表述为静态目标，本文对“胖系统”预测控制中静态目标的实现提出了综合性能指标法，改进了多目标优化法．

英文摘要

      One of the major assets of predictive controls lies in their ability to handle the non-square system through the performance criteria. The control of the "thin system" can be obtained directly in a "least-square"sense. For the "fat system", the controller has several degrees of freedom, which can be used to fulfill some economic Performance. In the paper, two methods of steady state object realization, the aggregative performance method and the multiobjective optimization method, are presented to meet the extra demands of the "fat system".