Evaluation of the Markov Technique with 20 Condition Status in Pavement Management System

The Markov technique with discrete interval and dividing the PCI into 10 modes and each mode into 10 units, gives the probability of the remaining in a specified mode or declining to the next mode after one duty cycle for the pavement. In this method can change the number of considered modes and extend the method accordingly. By increasing the number of modes, it may be possible to provide a more accurate estimate of the deterioration process and the type of actions needed at the appropriate time to increase road maintenance and preserve capital; Therefore, the purpose of this study is to investigate the division of PCI into 20 modes with 5 units’ width to estimate the pavement collapse process. In addition, the 5-unit segmentation is more in line with the presented qualitative classification. For this purpose, first, the probability matrix of the corresponding transfer is formed by two classifications, and then by a state vector and one family, a pavement performance prediction model based on homogeneous Markovian trend for two 10-digit and 5-digit PCI classifications, based on the data from valid twenty-year-old pavement was developed. The results showed that in the prediction curve obtained from the two segmentations, there is no specific critical point. But given a steep slope to begin to increase the severity of the failures, the 5-unit segment, 4 years earlier than the 10-unit segment, suggests the start of preventive measures. Also, the PCI prediction in the tenth year of these two curves shows the probability of 0.32 and 0.08 for the 10 and 5 unit methods, respectively, with significant differences. In addition, in the 5-unit method the curve slope increases over time, which is more in line with the pavement deterioration process, but in the 10-unit method the slope is almost uniform during the deterioration.