参数在线跟踪的直流传动双神经网络模型参考自适应控制

摘要

According to the limitation of the neural network controller in the pa per ‘Hopfield neural network (HNN) based model reference adaptive control (MRAC) of DC drive system’［8］when the load torque is unknown or time-varying, the param eter tracking scheme［2］ is introduced into DC drive system. Dual neural network based MRAC control pattern which has the ability of online parameters' tracking is constructed and the control performanc e is largely improved.%针对在“基于Hopfield神经网络的直流传动系统模型参考自适应控制”［8］一文 中指出的所提神经网络控制器在负载转矩未知或变化情况下的缺陷，将作者先前所提出的交流传动系统参数自动跟踪神经网络［2］引入直流传动系统，构成具有参数在线跟踪功 能的直流传动双神经网络模型参考自适应控制模式，进一步提高了系统的控制性能。