Abstract: In the case of false alarms, the inventory allocation of repairable spare parts is overestimated, and the inventory allocation of each station deviates from the demand. Combining the Metric theory, the characteristics of the two-echelon repairable spare parts support process are analyzed. The false alarm fault identification coefficient in the maintenance process is introduced to modify the average demand for spare parts of at each support station. Considering the cost of fault maintenance and false alarm identification, and satisfying the requirements of system performance indicators, an inventory allocation model aiming to minimizing the support cost is established. The marginal optimization algorithm is used for numerical analysis. By comparing the spare parts inventory allocation strategies with and without false alarms, it is proved that the model considering false alarms can improve the estimation accuracy of the two-echelon repairable spare parts inventory allocation.

Key words: system availability, false alarms, two-echelon inventory, marginal optimization