多变量解耦鲁棒控制的遗传优化方法

提出一种基于遗传算法的鲁棒准优势化算法。该算法将鲁棒对角优势和遗传算法相结合，综合考虑各摄动系统的关联性，搜寻最优的鲁棒解耦控制器。用该算法对一参数不确定性工业对象进行了鲁棒系统设计，结果表明该设计方法比多模型加权准优势化算法具有更好的鲁棒稳定性和鲁棒性能，因而更适于实际工程应用。

Multivariable Decoupling Robust Control Based on Genetic Algorithm

This paper presents an algorithm of robust pseudo-dominance based on genetic algorithm. Combining robust diagonal dominance with genetic algorithm, the method integrates the interactions of perturbed systems to search for the optimal robust decoupling controller. It has been applied to the robust system design for an industrial plant with parameter uncertainty. The results show that the design method exhibits better robust stability and robust properties than the multimodel-weighted pseudo-dominance algorithm,thus it is more useful in practical engineering applications.