Application of Markov process to pavement management systems at the network level

A probabilistic network level pavement management system (PMS) has been developed based on the Markov process. The system uses PCI method for pavement condition rating. The pavement performance prediction model has been developed based on the Markov process. Homogeneous and non-homogeneous Markov chains have been used in the model development. The comparison of the results from the Markov prediction model and the constrained least squares model showed similar trends. Dynamic programming has been used to develop the optimal M & R recommendations for each pavement family/state combination. Two prioritization programs have been developed to allocate the constrained budget in an optimal way. The first prioritization program is based of weighted optimal benefit/cost ratios. The second prioritization program is based on the incremental benefit/cost ratio technique.The implementation of the developed optimization methods on an existing airfield pavement network revealed that for a given constrained budget scenario: (1) Prioritization using incremental benefit/cost ratio results in a higher network PCI as compared to prioritization using optimal benefit/cost ratio. (2) The prioritization using the incremental benefit/cost ratio program utilizes the available constrained budget to the best of their full limit. (3) To maintain a specified network PCI, the incremental benefit/cost ratio program will spend more money as compared to the optimal benefit/cost ratio program.U of I OnlyETDs are only available to UIUC Users without author permissio