中文摘要：

目前火电机组优化控制系统中绝大多数属于开环优化控制方式,分析了这种方式的不足并建议采用闭环优化控制方式。极值搜索控制是一种传统的闭环优化控制方法,但存在需外加颤动信号和难于处理约束问题等不足,为此将极值搜索控制的思想和预测控制相结合,提出一种新的闭环优化控制方法—极值搜索预测控制,并对实现极值搜索预测控制需解决的关键问题和研究思路进行了讨论。

英文摘要：

Up to now,most of the optimization control systems in thermal power plants adopt the open loop manner.The drawbacks of such manner were analyzed and the closed-loop optimization manner was suggested in this paper.The extremum-seeking control（ESC） is a traditional closed-loop optimization control method.However,it requires a persistent dither signal exerting on the system and cannot deal with constraints effectively.To overcome these drawbacks of ESC,a novel approach-extremum-seeking predictive control（ESPC） was proposed and some essential problems and ideas were discussed to achieve ESC.