一种新型的智能预估器研究

RESEARCH ON A NOVEL INTELLIGENT PREDICTOR

本文讨论了一种新的基于特征状态的智能预估器。通过对典型化工大纯滞后对象的特性分析，提出了一种从对象控制作用和响应曲线中提取信息，构成特征状态，由此来驱动模糊规则的智能预估策略。它避开了传统的建模预估方式，更适于工程实际，其有效性通过实验得到了证实。

A novel intelligent predictor based on characteristic state is presented inthis paper. Through analysing the typical chemical process with large dead-time in details,a predictive strategy has been developed which extracts the process characteristic informationfrom its runing curve, use them to excite the fuzzy predictive rules, then get predictive value. The method avoids traditional model-based predictive manner. So it is more practical toengineering application. Experimental results show its effect.