Binary state systems can have two states either fully functional or totally failed. Multi-state systems (MSS) on the other hand can have more than two states such as functioning at a level lower than the installed capacity such as in a power generation plant or, one function of system alone not functioning as in a computer net work etc. The analysis of the system reliability becomes more complex as the number of parts increases. Including redundant sub-systems is one approach which is very widely used. However this adds to the cost and other parameters such as weight of the systems. So an optimal method of redundancy is the solution to this problem. This paper proposes an ant colony optimization (ACO) algorithm (Ref. 1) to minimize the total cost of the design in the redundancy allocation problem in a multi-state series-parallel reliability structures. A algorithm referred as SASACS-Algorithm, is presented here, to compute the reliability index of a system with maximum three types of components in each subsystem. (50 refs.)
展开▼