Predictive Controller Design For Coke Oven Blowing Cooler System Based On RBF

The coke oven cold drum system plays an important role in the coking process.It is mainly used to adjust the suction before the primary cooler and to realize the stable transmission of waste gas under different working conditions.Once the suction force before the primary cooler changes,the collector pressure will also become unstable.Due to the characteristics of multivariable,nonlinear,strong interference and strong coupling,the cold drum system cannot obtain accurate mathematical model by conventional modeling.Therefore,it is of great theoretical and practical significance to find a suitable modeling method to obtain a model with high accuracy and adopt advanced control technology to control the coke oven cold drum for a long time.Based on the reconstruction of a coking plant project,this paper divided the coal gas production into three working conditions: maintenance and heat preservation,normal production and coking and adding coal.In this paper,the model of coke oven cold drum system is identified by using the neural network RBF,and the identified model is combined with the model predictive control to form the predictive control based on RBF neural network,which is applied to the cold drum system control scheme.After the transformation of the coke oven cold drum system through the acquisition of field data,and the use of MATLAB software related simulation experiments,the experimental results show that the transformation of the cold drum system can be very good before the initial cooler suction stability in a certain range,fully meet the requirements of coking process.Although the suction before the primary cooler will fluctuate when it is disturbed at the beginning,it will stabilize quickly after a short period of time and return to the range required by the process.After the modification of the control scheme,the coke oven cold drum system has the characteristics of rapidity and anti-interference.