Fault tolerant system design optimization

A systematic design optimization methodology for fault-tolerant systems is presented. The mathematical formulation of the design optimization process exploits the conceptual analogy to the general optimal control problem. The fault-tolerant system dynamics is modeled and simulated using Markov models. Thus, unlike other approaches proposed in the literature, the proposed approach does not require explicit formulas for state probabilities and therefore is not limited to simple systems. The resulting formulation is a minimization problem for a cost function of interest. This optimization problem is solved using either a direct search or gradient-based techniques, to produce the set of design variables yielding the minimum cost value. The sensitivities of the cost function to the design variables are efficiently calculated as a natural result of the optimization routines. Two examples of the application of the proposed methodology are presented: a dual redundant warning lamp system and a hypothetical triplex redundant system