模糊神经网络在炼钢加热炉温度控制中的应用

Application of Fuzzy Neural Network in Steelmaking Heater Temperature Control

针对炼钢过程中的加热炉具有非线性、大滞后和建立精确数学模型有一定难度等特点,采用常规PID控制效果不是很好,设计了一种模糊神经网络自学习控制器,该控制器将离线训练和在线自学习结合起来。MATLAB仿真表明,模糊神经网络自学习控制器与PID控制相比较,改善了系统的动态性能,取得了良好的控制效果,达到了节能和环保的目的。

Based on steelmaking heater with nonlinear,large lagging and inability to establish accurate mathematical model,the traditional PID control is difficult to achieve the ideal control effect.A fuzzy neural network self-learning controller is designed.The controller combines the off-line training with on-line self-learning.Firstly,neural network parameters are trained off-line.Then the further online learning is performed.MATLAB simulation shows that the fuzzy neural network self-learning controller improves the dynamic and stable performance compared with PID control.The good control effect is achieved.The energy saving and environmental protection are got.