Applications Of Stochastic Process Models In Vehicle Fleet Management

Many enterprises or department have their own vehicle fleet. Vehicle fleet's effective management can improve its operation reliability and efficiency. Vehicle fleet management contains operation management and maintenance management. This thesis focuses on the later.In order to implement a maintenance management of vehicle fleet, it must be have an accurate description of vehicle fleet's failure process. Vehicle fleet is a complex system which contains many repairable subsystem and non-repairable subsystem. We should consider both situations when describe its failure process. Repairable system's reliability is no longer an inherent reliability; it is impacted by different maintenance and operation conditions. We should analysis repairable system's reliability based on reliability theory of repairable system rather than the traditional reliability theory. The stochastic failure point process model is the basic model in reliability theory of repairable system.This thesis has considered the application of two stochastic process models in vehicle fleet. One is the renewal process, which has a widespread application in varies areas. It regards as a renewal process while the elements of vehicle can't repair. It attached much focus in vehicle fleet's maintenance management. However, its application is restricted by the renewal function's formula's complex. This thesis proposes a new algorithm method to improve renewal function's computation. The new method is easy to realized and accurate. It is also suitable for other distributions. Then modeled base on a group failure data from vehicle fuel filter, and do some calculations use the new method. The number of fuel filter of a vehicle fleet during one year is forecasted. This result can give some help to purchase and stock department's work.Another stochastic process is the non- homogeneous Poisson process. We use power-law model modeled a failure process of a vehicle fleet, then optimization overhaul decision through it. This result provides useful advice for make reasonable overhaul decision system.