PIDNN Based Intelligent Control of Ignition Oven
Yuqian Ying,Jian'gang Lu,Jinshui Chen,Youxian Sun
DOI: https://doi.org/10.4028/www.scientific.net/amr.396-398.493
2011-01-01
Advanced Materials Research
Abstract:In a steel plant, fuel gas caloricity of ignition oven always changes rapidly and largely. Consequently, the temperature of ignition oven can’t keep steady. To overcome this problem we employ intelligent control of ignition oven based on PIDNN (Proportional-Integral-Derivative Neural Network). As we know, ignition oven is a nonlinear, large delay and slow time-varying process, so traditional PID control usually doesn’t work well. Artificial neural networks can perform adaptive control by learning, so we adopt Proportional-Integral-Derivative neural network to tackle the problem taking the advantages of both PID control and neural structure. In order to satisfy the restrictions of industrial instruments, we combine PIDNN control algorithm with expert system mechanism to fulfill the final intelligent control strategy. At a sintering plant in Hangzhou, we deploy the intelligent control strategy turning out a satisfactory result that the ignition oven temperature can be controlled steadily within a much smaller range with significant saving of labor costs and improving of energy efficiency.