Application of neural networks to heat transfer calculation in a round billet continuous casting mould
-
摘要: 以圓坯連鑄結晶器溫度實測數據為基礎,研究應用神經元網絡求解結晶器傳熱反問題的可行性,并嘗試將神經元網絡與數值計算結合用于結晶器傳熱計算,探索針對實測數據的結晶器傳熱在線計算方法.結果表明,數值計算結果與實測溫度符合較好,可真實反映實際生產中圓坯結晶器傳熱的非均勻特性.Abstract: Based on the measured data of mould temperature during continuous casting of round billets, the feasibility of using neural networks to solve the inverse problem of heat transfer of a continuous casting mould was discussed. The combination of neural networks with numerical simulation was proposed to calculate the heat transfer of the mould. The results show that the temperature and heat flux field calculated are in agreement with the experimental ones, which could correctly reveal the characteristics of non-uniform heat transfer around the mould circumference.
-
Key words:
- round billet /
- mould /
- heat transfer /
- neural networks /
- inverse problem
-
點擊查看大圖
計量
- 文章訪問數: 129
- HTML全文瀏覽量: 21
- PDF下載量: 5
- 被引次數: 0
下載:
