Availability Optimization for Series/ Parallel Systems using Evolutionary Algorithm

One of the best solutions provided to improve the availability of industrial and communication systems is to add redundant elements. Usually this task tries to achieve a performance level with minimal cost. The problem of total investment cost minimization, subject to availability constraints, is a pure redundancy optimization problem which is known to be intractable and for which no deterministic algorithm is expected to succeed. The difficulties come from the constraints and variable domains which are non-linear and of mixed types. To tackle the redundancy optimization problem for a system with different capacities, we propose two problem-specific evolutionary algorithms for the cost minimization and for the availability maximization. For the first case, the algorithm makes use of evolutionary operators that map feasible points into other feasible points. Specialized operators for boundary search allow an optimal utilization of available resources, since the search is restricted to solutions corresponding to a full utilization of these resources. In the second case, boundary search operators cannot be explicitly stated. The proposed alternative is a repair algorithm that maps infeasible solutions to feasible ones as close as possible to the boundary of feasibility, if not exactly on it.