Point-to-point iterative learning control with quantised input signal and actuator faults

This paper applies iterative learning control to point-to-point tracking problems with a general networked structure. The data is quantised and transmitted through restricted communication channels from the controller to the actuator. Combining a logarithmic quantizer with an encoding and decoding mechanism to quantise the input signals reduces the influence of the quantisation error. New design algorithms are developed with conditions for convergence of the tracking error and an extension to fault-tolerant performance under actuator failures. A numerical-based case study demonstrates the application of the new designs, which includes a comparison with another ILC law and the relative merits of the encoding and decoding schemes.