Modeling Based On QPSO-BPNN And Control Of The Secondary Cooling Water In Continuous Casting

In the process of continuous casting, the slab with a long liquid heart goes through the secondary cooling zone (SCZ). It will be completely solidified after the secondary cooling spray water to cool. In this cooling process, the effect of the secondary cooling water (SCW) will directly affect the quality of the slab, so suitable control to SCW has a positive significance to the process of continuous casting.The control method of SCW is mainly divided into two categories:one is static control, and the other is dynamic control. Though widely used, but the disadvantage of static control is obvious. Thus we should vigorously develop dynamic control. One of dynamic control method is adjusting the waterflow according to the slab surface temperature of the slab. But due to the strong interference from the continuous casting SCZ, it is difficult to measure the surface temperature of slab. Therefore, the application of this method has been limited. The other method is module based, controlling the water flow by comparing the objective temperature of the surface of the slab with that predicted by a model.For the model based method, we should pay attention to the accuracy of the model as well as its practicability. At present, most dynamic control models are slab coagulation models, which are not practical since they need a lot of computing. So we build the temperature prediction model of continuous casting SCW by BP Neural Network (BPNN) optimized by Quantum Particle Swarm Optimization (QPSO).In this paper, we first start from the basic theory of SCW. Then an analysis of the advantages of QPSO and the disadvantages of BPNN is given. Based on this, we decide to use BPNN, optimized by QPSO, to build the predictive model of cast SCW. Then we give a simulation of the proposed prediction model.Finally, based on this model we establish the continuous casting SCW control system using a self-adaptive PID controller. The results of simulation prove that the control system can provide a good performance.