配煤过程的神经网络专家控制策略

Neural Network Expert Control Strategy for Coal Blending Process

根据钢铁工业的配煤过程 ,提出一种精确地决定和跟踪配煤比的神经网络专家控制策略 .首先基于过程的统计数据和经验知识 ,建立了神经网络、经验模型和规则模型 ,然后结合这些网络和模型 ,提出一种决定配煤比的专家推理方法 ,最后用分布式控制系统实现配煤比的跟踪控制 .

A neural network expert control strategy was proposed for the coal blending process in the iron and steel industry. It is used to accurately determine the target percentage of each type of coal to be blended, and to track them. In this paper, neural networks, empirical models and rule models were first constructed based on statistical data and empirical knowledge. Then, an expert reasoning method combining the networks and models was proposed for determining the target percentages. Finally, a distributed control system was used for tracking the target percentages. The results of actual operation show the validity of the strategy.