A Combustion Control Modeling of Coke Oven by Swarm-based Fuzzy System

This paper proposes a swarm-based fuzzy system modeling technique for coke oven combustion control diagnosis. Coke is produced by the coke plant with charging coal into oven and supplying gas to carbonize it and it is used at the blast furnace plant in the steel making process. A conventional mathematical model for coke oven combustion control has been used to control the amount of gas input, but it does not work well because of highly nonlinear feature of coke plant. To solve this problem, swarm-based fuzzy system modeling technique is suggested to construct a diagnosis model of coke oven combustion control. Based on the measured input-output data pairs in the plant, the fuzzy rules are generated and the parameters are tuned by the particle swarm optimizer to increase the accuracy of the fuzzy system is operated. This system computes the proper amount of gas input taking the operation conditions of coke oven into account, and compares the computed result with the supplied gas input. The result of simulation shows that developed fuzzy system can be used as a diagnostic tool for coke oven combustion