改进ELMAN网络的Q学习温度切换控制

Temperature switching control integrated with improved ELMAN network and Q learning

中文关键词  料分解过程  分解炉温度  改进ELMAN网络  Q学习控制  切换控制

英文关键词  raw meal calcination process  calciner temperature  improved ELMAN network  Q learning control  switching control

中文摘要

      在水泥生料分解过程中, 由于生料边界条件频繁变化且生料流量波动大, 使过程处于正常工况或异常工况, 致使分解炉温度和预热器C5出口温度很 难控制在工艺要求的范围内. 传统的控制方法经常导致预热器C5下料管堵塞. 为了解决上述问题, 本文提出了一个带有前馈补偿的温度智能切换控制策略, 由基于T--S的模糊控制器、基于改进的ELMAN网络的Q学习异常工况控制器和切换机制组成. 实际应用结果表明, 所提出的控制策略能够根据当前工况的变化选择正确的控制器, 并且使生产远离故障工况.

英文摘要

      Because of the frequent change of the boundary conditions as well as the large variation in the law meal flow in calcination process, the process is often switching between the normal and abnormal working conditions. This causes difficulties to maintain the calciner temperature and the outlet temperature of preheater C5 (i.e., the No. 5 preheater) within their desired range of operations. Conventional control may even lead to operational failure, such as the clogging of feeding tube C5. To deal with this problem, we develop an intelligent temperature control scheme with feedforward compensation, which consists of a T–S-based fuzzy controller, an abnormal condition controller, an improved ELMAN network with Q learning, and a switching mechanism. The application results show that the proposed approach can select a right controller based on the current conditions to maintain the operation far from the faulty operation conditions.