基于神经网络逆复合控制器的MDF施胶系统的控制设计

根据神经网络逆控制原理,提出了一种新的中密度纤维板(MDF)施胶系统控制器设计方法。将MDF施胶系统作为被控对象,利用静态神经网络和若干积分器组成的动态神经网络构造出被控系统的逆系统,该逆系统与被控系统串接在一起所组成的系统是伪线性的。通过将PID控制器与神经网络逆系统结合在一起,构建出神经网络逆复合控制器。仿真结果表明:神经网络逆复合控制器能使系统获得优良的跟踪性和抗干扰能力,可以很好地实现对MDF施胶系统的控制目的。

Control Design of MDF Glue System Based on Neural Network Inverse Composite Controller

A new controller designed method applied to the Medium Density Fiberboard(MDF) glue system was proposed based on neural network inverse control theory.And an inverse system of the controlled system was built with a dynamic neural network composed of static neural network and several integrators using the MDF glue system as the controlled system.PID controller was combined with neural network inverse system to build a neural network inverse composite controller due to the pseudo-linear series system composed of the inverse system and the controlled system.The simulation results show that the neural network inverse composite controller can make the system gain excellent tracking performance and anti-interference ability;thereby an anticipated control purpose of the MDF glue system can be obtained.