基于遺傳神經網絡的冷連軋機軋制壓力模型

Rolling Force Models of Cold Tandem Rolling Mill Based on Genetic Neural Networks

摘要: 對鞍鋼冷軋廠四機架冷連軋機軋制壓力模型進行了認真分析,指出了其存在的缺陷.把遺傳算法(GeneticAlgorithms,簡稱GA)和神經網絡有機結合,設計出了具有遺傳算法性能參數優選、網絡結構參數優選、網絡性能參數優選以及GA-BP算法聯合進行網絡權值修改幾種功能的遺傳神經網絡,建立了基于遺傳神經網絡的新沖連軋機軋制壓力模型.通過原模型計算值、新模型計算值與實測值之間的對比分析可知,遺傳神經網絡模型計算精度優于傳統軋制力模型.

Abstract: Some defects of the traditional rolling force models of cold tandem rolling mill were found out, and new rolling force models based on genetic neural networks were set up. The comparison results of the measured rolling force of cold tandem rolling mill with the calculated value from the traditional models and also with the calculated value from the new rolling force models based on genetic neural netWothe show thatthe calculating precision of the new models is better than that of the traditional models.