基于NN-IMC模型的结晶器振动波形控制研究

Mold Vibration Waveform Control Research Based on NN-IMC Model

根据神经网络系统和内模控制系统的基本原理,结合结晶器振动控制结构,建立神经网络内模控制(NN-IMC)系统。通过分析,采用范化回归神经网络GRNN进行系统控制,并对连铸连轧结晶器振动波形进行系统仿真。仿真结果表明NN-IMC系统能够很好地跟踪结晶器振动波形,满足系统的性能要求。

According to the basic principles of the neural network system and the internal model system, combined mold vibration control structure, the neural network-internal model control （ NN-IMC） system is established. System control is performed with generalized regression neural network（ GRNN） , and mold vibration waveform of continuous casting and rolling is simulated. The simulation result indicated that the NN-IMC system could track mold vibration waveform well, and satisfy performance requirement.