Iterative Learning Control for Encoding-Decoding Method with Data Dropout at Both Measurement and Actuator Sides

This paper investigates the problem of iterative learning control in networked structures, with a specific focus on addressing random data dropout at both measurement and actuator sides to achieve zero-error tracking performance. An encoding and decoding mechanism is introduced into the system. Initially, the system output undergoes encoding, quantization, and subsequent transmission to the controller. When the data are received, they are decoded and applied to generate the input for the next iteration. Subsequently, the generated input undergoes the same process as the output transmission, including encoding, quantization, transmission, and decoding. The proposed approach's convergence is proven under random data dropout, and the effectiveness of the scheme is demonstrated through convergence analysis and illustrative examples.