论文部分内容阅读
电极电流值和极心圆直径是电熔镁砂生产过程中影响产品性能指标的两个重要参数,在传统工艺中由人工设定.由于生产过程中存在很多复杂特性,人工给定的设定值很难保证其准确性.针对这种情况,提出了基于案例推理、迭代学习、PI控制、神经网络和规则推理的参数混合智能设定方法.该方法成功应用于国内某电熔镁砂厂,应用效果表明了所提出方法的有效性.
The electrode current value and the diameter of the perfect circle are two important parameters affecting the product performance in the production of fused magnesia, which are manually set in the traditional process. Due to the complicated features in the production process, the artificial setting It is difficult to guarantee its accuracy.According to this situation, a hybrid parameter setting method based on case-based reasoning, iterative learning, PI control, neural network and rule reasoning is proposed.The method is successfully applied to a fused magnesia plant in China The application effect shows the effectiveness of the proposed method.