一种智能状态估计方法及应用

为提高机理复杂工况变化较大的一类生产过程状态估计的精度 ,提出一种基于多模型和专家知识的智能状态估计方法 ,并给出了用于青霉素发酵过程状态估计的结果。这种方法采用多个状态估计模型进行估计 ,然后由优选专家系统选择一个最优的作为最终结果。给出了状态估计系统的结构及优选专家系统的结构和设计。优选专家系统采用改进的 Bayes方法进行不确定性推理。这一估计方法可以应用于类似于青霉素发酵这类机理复杂过程的状态估计。

Intelligent state estimation method and its application

The precision of state estimates for production processes that have complicated mechanisms and hence great variety is improved using an intelligent state estimation method based on multi models and expert knowledge. The method is used to the estimate the result for the penicillin fermentation process. The method uses multiple process models to estimate process states and an optimal seeking expert system to select the optimum state as the final estimated result.The paper gives the framework for the estimation system and the structure of the optimum seeking expert system .The optimal seeking expert system uses the improved Bayes method for the uncertain reasoning. This method can be used to estimate states for various complicated processes such as penicillin fermentation.