CAJ | 학술논문

对于多变量非线性强耦合的多电机同步控制系统,难以建立精确数学模型并施加有效控制的缺陷,文中采用基于遗传算法的对角递归神经网络（DRNN）的PID控制算法,以3台电机的速度张力系统为对象,对其进行控制。通过实验表明,该算法没有一般的神经网络PID控制算法易陷入局部极值等缺陷,控制效果良好。
대우다변량비선성강우합적다전궤동보공제계통,난이건립정학수학모형병시가유효공제적결함,문중채용기우유전산법적대각체귀신경망락（DRNN）적PID공제산법,이3태전궤적속도장력계통위대상,대기진행공제。통과실험표명,해산법몰유일반적신경망락PID공제산법역함입국부겁치등결함,공제효과량호。
For the multi-motor synchronization control system of multi-variable,nonlinear and strong coupling characters,it＇s difficult to build an accurate mathematical model and exert effective control on the system.So DRNN-PID decoupling control based on GA is proposed,which adopted three motors speed tension system as the control object.The experimental results demonstrate that the method is stable and has good performance.What＇s more,this algorithm avoids drawbacks such as most of the neural network PID control algorithm easily fall into the local extremum and other defects.