摘要
针对连铸传热模型参数辨识问题中包含偏微分方程所带来的复杂性和非线性,提出采用混沌粒子群算法进行优化求解,依据不同位置射钉测量坯壳厚度和二冷外测量铸坯表面温度相结合,优化确定了二冷换热系数和有效导热系数相应参数.最后通过在线计算铸坯表面温度与二冷出口铸坯表面测温比较,结果偏差小于13℃,验证了辨识参数的可靠性.校验后的模型成功应用于连铸机的二冷配水优化和动态控制.
Due to the complexity and nonlinearity with a partial differential equation, an identification method based on chaos particle swarm optimization (CPSO) was presented. Based on the combination of the measured shell-thicknesses by nail-shooting at different positions and the measured surface temperatures outside the secondary cooling chamber, parameters related to the heat transfer coefficients of secondary cooling and the effective conductivity were determined. Finally the model was verified as the difference was within the range of ± 13 ℃ between the calculated surface temperature by online calculation and the measurements at the exit of secondary cooling chamber. The verified model was applied to the optimization and dynamic control of secondary cooling.
出处
《东北大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2014年第5期613-616,共4页
Journal of Northeastern University(Natural Science)
基金
国家自然科学基金资助项目(61273178)
关键词
连铸
传热模型
参数辨识
粒子群算法
混沌
continuous casting
heat transfer model
parameter identification
particle swarmoptimization
chaos