An Overview of Pavement Degradation Prediction Models

Table 1

Pavement performance prediction studies in probabilistic reasoning.

Technique used	Pavement indicator	Data sources	Metrics	References	Strength	Weakness
Markovian model	IRI	Observed		Porras-Alvarado et al. [40]	Capable of evaluating multiple hypotheses and accurate with first-order Markov property	Not suitable in case of higher-order correlation and missing data.
		LTPP	RMSE	Alimoradi et al. [41]
Empirical mechanistic	PCR	Observed		George et al. [23]	It contains features from statistical and mathematical elements	Lake sights for pavement preservation and limited literature.
Fuzzy logic	IRI	LTTP	RMSE and RE	Li [33]; Nguyen, et al. [32]	Simple control, more robustness, and more efficiency with control systems	Have steady-state errors, weak in real-time response, and a limited number of input variables
Deterministic models (Al Omari–Darter model and Dubai model)	IRI	Observed	R^2	Al-Suleiman and Shiyab [34]	Good in decision making by providing clear information of the future trends and challenges.	Easy to misinterpret and hard to check the validity.
		LTTP	R^2	Chen and Zhang . [37]

R²: coefficient of determination, RMSE: root mean squared error, MAPE: mean absolute presenting error, CF: correction factor, VAF: variance account for, MAE: mean absolute error, RE: relative error, MSE: mean squared error, SDMSE: standard deviation of mean squared error.