Anti-disturbance iterative learning tracking control for general non-Gaussian stochastic systems

In this paper, a class of general non-Gaussian stochastic systems with disturbances are studied. Based on the disturbance observer (DO) design method, an anti-disturbance iterative learning control (ILC) algorithm is proposed by establishing the statistic information tracking control (SITC) framework. Different from the existing stochastic control methods, the driven information for control feedback is the output statistic information sets (SISs) relying on sample data of the non-Gaussian stochastic output, rather than the output PDFs. A novel model-free ILC optimization problem is addressed by combining the DO design with ILC algorithm. The controller design can be achieved based on the convex optimization to ensure the configured system stability and convergence of the tracking error to zero. Meanwhile, the satisfactory disturbance estimation and rejection performance can also be guaranteed. In the simulation, a typical 3-parameter Weibull distribution is considered to demonstrate the effectiveness and the practical significance of the proposed algorithm.