Probabilistic optimization of aging structures considering maintenance and failure costs

The prioritization of maintenance interventions on aging structures is one of the most important decision processes in the management of existing civil infrastructure. This process has to consider the uncertainties in loads, resistances, and maintenance interventions over a given time horizon. A computational program for probabilistic maintenance optimization in which the reliability index is the quantitative measure of performance in the presence of uncertainties and the present value of expected cumulative cost is the objective to be minimized has been developed by the writers. Using this program, time-varying uncertainties associated with reliability deterioration of structures and effects of maintenance interventions on the system reliability can be captured efficiently. The aim of this study is to provide insight in the methodology for probabilistic maintenance optimization of aging structures in which the optimum solution is found by minimizing the present value of expected cumulative maintenance and failure costs over a prescribed time horizon. Because of the discreteness in cost profile evaluation and randomness in the feasible domain, the conventional optimization methods are modified by using the branching technique and multidimensional interpolation method. Both time- and reliability-controlled maintenance actions including essential and/or preventive interventions on aging structures are optimized.