基于自适应蚁群算法的传热模型参数辨识

来源期刊：冶金自动化2009年第1期

论文作者：纪振平 谢植 马交成

文章页码：6 - 11

关键词：连铸； 蚁群优化； 传热模型； 参数辨识

Key words：continuous casting； ant colony optimization； heat transfer model； parameter identification

摘    要：针对铸坯凝固传热模型校正问题,提出了采用自适应蚁群优化算法进行参数辨识的方法。首先在不同条件下对铸坯不同位置进行射钉并测量凝固坯壳厚度,然后根据测量数据集和凝固传热数学模型的数值解,采用自适应蚁群优化算法进行参数辨识,确定二冷区内各冷却段的传热系数,最后通过二冷出口铸坯表面温度和凝固坯壳厚度的测量数据与采用辨识参数的传热模型预测结果进行比较,验证了传热系数的准确性。校正的传热模型已成功应用于多台铸机的二冷配水优化。

Abstract: For calibration of solidification and heat transfer model of billet,a parameter identification method based on adaptive ant colony optimization(AACO) algorithm was developed.Firstly,solidification shell thickness of billet was measured by nail shooting at different positions under different conditions.Following that,according to measured data set and numerical solution of solidification heat transfer model,parameter identification was carried out through adaptive ant colony optimization algorithm and heat transfer coefficient at any cooling segment of secondary cooling zone was determined.Finally,measuring temperature at exit of secondary cooling zone and shell thickness of billet were compared with calculation results of solidification heat transfer model with identified parameters.The results show that determined parameters are correct.The calibrated model has been used in several casters for optimization of distribution of secondary cooling water.