This paper enhances a recently developed tool to analyze the mean-square stability of digital control systems implemented in computers or embedded processors with rollback recovery that are operating in harsh environments. In particular, closed-loop models are derived that allow the analysis of partial state rollback where not all the states are check-pointed and rolled back. The theoretical results are illustrated with the analysis of a stabilizing controller for the longitudinal dynamics of the AFTI/F-16 aircraft. This example is also used to determine the effect of longer checkpoint durations on the mean-square stability of the flight control system.
展开▼