粗糙集理论在简化过程建模参数中的应用

利用粗糙集理论知识约简的特点，以连续催化重整反应过程中重整产品辛烷值监控建模参数选择为例，对生产过程工况特征参数建立决策表，计算条件属性的重要性，对决策表进行属性约简，简化过程建模参数。此外还将该方法与主成分分析法作了比较。最后利用选择的建模参数建立起神经网络模型。

Application of Rough Sets Theory in the Parameter Simplification of Process Model

This paper takes advantages of the rough sets theory in knowledge-reduction, then takes octane supervision of CCR as an example, builds up the decision table by the characterized parameters, calculates the significance of condition attributes. Unnecessary attributes are eliminated from the decision table after the attributes reduction. It achieves the purpose of parameter simplification of process model. After compared with the method of principal component analysis, it gives the Neural network model based on the selected parameters.