Method to improve convergence performance of iterative learning control systems over wireless networks in presence of channel noise

The accumulated effect of channel noise, including sensor-to-controller (SC) noise and controller-to-actuator (CA) noise, has a significant impact on the convergence performance of iterative learning control (ILC) systems over wireless networks. In this study, the relation between input error, channel noise and learning gain is derived, which reveals the fact that the contribution of the SC noise and the CA noise to the input error are all influenced by the learning gain. Based on this discovery, a method is proposed to improve the convergence performance of the ILC system when the SC noise and CA noise are independent and Gaussian distributed. Specifically, this method adaptively selects the learning gain through minimising the trace of input error covariance matrix. With the adaptively selected learning gain, the convergence performance of the ILC system is improved significantly. Moreover, the effect of channel noise variance on the convergence speed of the ILC system with the proposed method is analysed theoretically. Finally, numerical experiments are given to illustrate the effectiveness of the proposed method and corroborate the theoretical analysis, respectively.