Research Of Predictive Control In Coke Oven Temperature Control

Coke is the main raw material and fuel of metallurgy, machinery and chemical industry, at the same time our country is the biggest coke producer and exporter in the world. The coke oven is the most complicated thermal technology equipment in the thermal kiln, it has characters of strong time delay, great inertia, strong non-linear, multi-factor coupling and variable parameter. The stability of coke oven's temperature has a very important meaning in reducing the fuel consumption of coke oven, improving the quality of coke and lengthening the longevity of oven.In this article, on the background of the NO.1 coke ovens of Benxi iron factory, on the basis of consulting a large number of domestic and international relevant documents, the development of coke oven's heating system is proposed firstly. Then through analyzing coke oven in depth and investigating generalized predictive control and dynamic matrix control algorithm, predictive model of implicit generalized predictive self-tuning is set up.The algorithm inherited the advantages of traditional generalized predictive control algorithm on one hand; on the other hand, error correction method of dynamic matrix algorithm is introduced into the algorithm. The algorithm uses error prediction to supplement the prediction based on model, namely other prediction exported in moment of future is revised through the weighted prediction error model. This algorithm does not need to distinguish target's model parameter, but according to distinguishing directly to get the parameter of optimum control rule. So a large amount of computation for calculating the Diophantine equation is avoided, time is saved. Applying this algorithm to the coke oven's temperature model, through comparing with the simulation of PID control in static characteristic, dynamic characteristic, the ability of anti-disturbance, robustness, and satisfied control quality of this algorithm is proved.