论文部分内容阅读
分散控制是大系统理论中一种重要的控制结构,其主要特征表现为局部的决策只依据局部的信息。它在控制的实时性、经济性、可靠性方面带来了若干优点,但信息结构的限制常会导致控制时的信息不足,因而引起性能下降、稳定性变差等一系列问题。预测控制是过程领域内发展起来的一类新型计算机控制算法。与传统的最优控制相比,它有两个新特点:一个是用投影预测补充模型预测的不足;二是采用滚动时域的启发式优化。这些使预测控制对模型的要求低,并能考虑模型中未能顾及的不确定因素,控制的鲁棒性好。
Decentralized control is an important control structure in large-scale system theory. The main characteristic of decentralized control is that local decisions are based on local information only. It brings some advantages in terms of real-time control, economy and reliability. However, the restriction of information structure often leads to the lack of information in control, which leads to a series of problems such as performance degradation and poor stability. Predictive control is a new type of computer control algorithm developed in process field. Compared with the traditional optimal control, it has two new features: one is to predict the lack of supplementary model predictive projection; the other is to use heuristic optimization of rolling time domain. These make predictive control less demanding on the model, and can take into account the uncertainties that can not be taken into account in the model, and the robustness of the control is good.