论文部分内容阅读
以缩小连铸二冷区板坯表面实际温度和目标温度的差异为目标,建立了板坯连铸二次冷却智能控制模型.该模型采用支持向量机(SVM)实现板坯表面目标温度的动态设定,采用对角递归神经网络(DRNN)实现板坯表面温度的预测,采用T-S模糊递归神经网络实现二次冷却水动态调整与分配.通过对某钢厂板坯连铸过程进行仿真计算和现场试验,结果表明:该模型将二次冷却水水量控制问题与板坯在冷却过程中的温度状态相结合,实现了连铸二次冷却动态优化控制,有利于提高板坯的质量.
In order to reduce the difference between the actual temperature and the target temperature of slab surface in the second continuous cooling zone, an intelligent control model of secondary cooling of slab continuous casting was established.The dynamic model of the slab surface temperature was established by using support vector machine (SVM) The diagonal recurrent neural network (DRNN) was used to predict the slab surface temperature, and the TS fuzzy recurrent neural network was used to adjust and distribute the secondary cooling water dynamically.Through simulating the slab continuous casting process in a steel plant and Field tests show that the model combines the control of secondary cooling water flow rate with the temperature of the slab in the cooling process, which realizes the dynamic optimal control of the secondary cooling of continuous casting and improves the slab quality.