基于RBF神经网络钢包烘烤装置的节能优化

Ladle Baking Devices Energy-saving optimization based on RBF neural network

以唐山钢铁股份有限公司第一、二钢轧厂的蓄热式钢包烘烤装置为控制对象,将基于RBF神经网络的PID控制方法与双交叉限幅方法结合引入到蓄热式钢包烘烤装置的温度控制中,实现了蓄热式钢包烘烤装置温度控制的节能优化。通过在Matlab环境下将RBF神经网络PID控制和传统的PID控制方法进行仿真研究,得出了最佳PID参数,将PID最佳参数应用到双交叉限幅控制方法中。仿真结果表明:基于RBF神经网络PID控制方法较传统的PID控制方法与双交叉限幅方法结合,其PID控制器快速性好,自适应力强,有更好的控制效果。并将该控制系统成功应用到唐山钢铁股份有限公司第一、二钢轧厂的蓄热式钢包烘烤装置温度控制中,在保证质量和产量的基础上,节省了混合煤气消耗量,使其节能32%。

This paper considers the regenerative ladle baking devices in First and Second Steel Rolling Plant of Tangshan Iron and Steel as control objects. We combine the method of PID control based on RBF neural network with double-crossing clipping and apply this combined method in temperature control of the regenerative ladle baking devices, so that the energy-saving optimization of those devices can be achieved. The optimal PID parameters are deduced through the simulation research of PID control based on RBF neural network and traditional PID control method in MATLAB environment and applied it to the double-crossing clipping control method. Then we demonstrate a simulation of the regenerative ladle baking devices application to indicate the high control by this combined. The PID con- troller is fast and good adaptive capability. And this control system has been successfully applied to regenerative ladle baking devices temperature control in the first and second steel roiling plant of Tangshan Iron and Steel, the energy-saving rate is 32 % and saving the mixing gas consumption on the basis of ensuring the quality and output.