A multi-agent predicting method of fleet spare part requirement for fleet condition based maintenance

To meet the demand for spare part predicting of a typical mission, a multi-agent based fleet spare part predicting method is proposed. The process of fleet spare part replacement is analysed, and the mathematical expression of fleet spare part predicting problem is introduced. Then the process is described as a 3-layer multi-agent model, and the fleet spare part replacing policy and predicting algorithm are proposed based on the model. Finally, a fleet of 8 aircrafts is studied for verification purpose, a mission type with 4 waves continuous dispatched is imaged. The proposed method can not only solve spare part demand, but provide optimal replacement strategies as well.