具有时延和数据丢失的直线电机迭代学习控制

针对网络环境下同时存在一步时延和数据丢失的直线电机系统,本文研究了其P型迭代学习控制方法.将一步时延描述为概率已知的随机Bernoulli过程,而将数据丢失描述为概率未知的随机Bernoulli过程,且丢失概率属于一个已知的数值区间.然后,利用λ范数理论给出了P型迭代学习控制算法的收敛条件,通过适当选取学习增益因子,可使得直线电机控制系统的输出渐近收敛于期望输出.最后通过仿真验证了本文所提控制方法的有效性。

Iterative learning control for linear motor system with time delay and data dropout

For linear motor control system based on networks with one-step delay and data dropout, a P-type iterative learning control method is studied in this paper. The one-step delay is described by a stochastic Bernoulli process with a known probability. The data dropout is also described by a stochastic Bernoulli process, whose probability is unknown and belongs to a known numerical interval. Then the convergence condition of the P-type algorithm is derived by using  norm theory. The output of the linear motor converges to the desired output asymptotically through choosing learning gain factor properly. Finally a simulation is conducted and the results demonstrate the validity of the proposed design method.