This study is concerned with a military activity operating fleet of multiple units essential for national defense in an operational base and repair depot combination. Analysis of the fleet operations is necessary not only to maintain effective national security but also to efficiently utilize the defense funds available.
Interactions of major factors in fleet operations are modeled in the framework of a closed queueing network system with a finite number of units assigned to the fleet. The factors affecting the fleet operations, such as unit``s characteristics and population, operational policies and constitution of repair depot including specialized repair facilities and spare module inventory, are intended to approximate the real situations more closely than other related works in the literature.
A recursive computing algorithm is developed to evaluate the effect of operational policies and investment decisions for repair depot supportability on fleet availability. Fleet availability is an important part of operational effectiveness for such fleet deployed at forward areas of conflict and can be measured as average fraction of units operable at a random instant.
In addition, this study provides a resource allocation program for improving the fleet availability when a limited amount of funds is available. The program focuses on allocating the funds available to multiple repair and inventory stations in the repair depot competing for their use. The optimization problem on hand is a mixed integer-nonlinear programming problem. Since an adequate programming algorithm has not yet been developed in solving the problem, heuristic methods based on gradient techniques are employed. Approximate solutions, although not optimal, can be obtained efficiently by the method.